Update note: this post went up on 9/9 and was updated through the week as the soap opera kept going: the letter from 25 Fields medalists (9/11), Dario Amodei’s essay and the response from Altman and Musk (9/12), Trump’s remarks (9/13), fact-check fixes raised in the comments, and Pirate Wires’ report on the PR firm behind Coxon (9/25).

This week was a festival of aggressive advertising from the AI companies. There were so many pompous declarations in so little time that you could use it as a case study in deceptive marketing. So let me repeat my usual motto, because it never fails:

“Your excitement about AI is inversely proportional to your knowledge about AI”

If you read this week’s headlines and walked away excited that the Terminator is coming, or that the machine just became God, you fell for the advertising. And I’m going to take it apart piece by piece so you understand why.

And let me put my cards on the table right away, because I’ve been saying this for a while on the podcasts I show up on. To me, Dario Amodei, from Anthropic, has a God complex. The guy truly believes he is saving humanity, with that creepy woke messianism of his. And Sam Altman, from OpenAI, is a mobster. He fancies himself Michael Corleone, the cold, calculating boss from The Godfather, but in practice he’s Fredo: the weak, bumbling brother who thinks he’s the smart one and keeps screwing up. Keep these two portraits in mind, because everything I’m about to explain below proves exactly that.

“AGI has arrived,” says the guy selling the shovel

It started with Jensen Huang, CEO of Nvidia, declaring once again that AGI has arrived. Not in a paper, not in a technical demonstration, but in a Sunday post on X, tying the declaration to OpenAI’s new Astra model, trained on roughly 100,000 Grace Blackwell chips. In other words, the chips Nvidia sells.

Look at the source. The man who makes money selling shovels to the gold miners announcing that the gold is infinite. Remember his appearance on the Lex Fridman podcast a few months back. Lex floated a tongue-in-cheek definition: AGI would be an AI able to start, grow, and run a tech company worth a billion dollars by itself. And Jensen, without blinking: “I think we’ve achieved AGI”. A conveniently commercial definition that has nothing to do with intelligence, and that the shovel salesman embraced on the spot. Gary Marcus went straight to the obvious question: does Jensen have any financial incentive to declare that AGI has arrived, given that he sells the GPUs?

And here’s the irony: despite the grandiose headline, Nvidia’s own stock dropped about 2% in the following days. Even the market, which runs on hype, didn’t fully buy the story.

Right after, OpenAI launched GPT-6 Astra and hammered the same speech. Greg Brockman closed the press conference with a “welcome to the AGI era” and said “for me, personally, I think we’re there.” Except, in the same breath, he dropped the line that kills the whole thing: “everyone has a different definition of AGI.” So the guy claims the most important thing in human history has arrived and, at the same time, admits nobody really knows what that thing means. That alone should set off the alarm in your head.

The “Millennium Problem” brute-forced

The main course of the propaganda was OpenAI announcing it had solved a math Millennium Prize Problem. Let me explain what that is, because the headline is built precisely to impress people who don’t know.

In 2000, the Clay Mathematics Institute listed seven open problems in mathematics and offered a million dollars to whoever solved each one. These are brutally hard problems. In over twenty years, only one had been solved: the Poincaré Conjecture, by Grigori Perelman, who, by the way, turned down the prize.

One of those seven is the Navier-Stokes equations. These are the equations that describe how fluids move: water, air, blood. The open question is not “how do you simulate fluid”: engineering has done that just fine for decades in any aerodynamics software. The question is purely mathematical: does a smooth solution always exist, or can the fluid develop a singularity, a point of infinite velocity, the so-called “blow-up”, in finite time? It’s a question about the rigor and existence of solutions, not about the practical usefulness of the equations.

And here’s my first point, which I already explained in detail in a post: even solving this definitively changes absolutely nothing in real life. No airplane flies better, no weather forecast gets more accurate, no medicine works any differently. Engineers keep modeling fluids the same way, with or without the proof. It’s a beautiful intellectual achievement, period. The practical impact is zero.

Now for the detail the headline hides. The announced proof attacks the forced version of the equations, where a smooth external force term is pushing the system, and it nails the blow-up. Technically it counts toward the prize, because those routes are part of Clay’s official statement. But the famous question, the deep heart of the problem, is the version without external forcing, and that one remains open. Clay has not closed the case, and OpenAI didn’t even go collect the million-dollar prize. (If you want the fine print on this, there’s an appendix at the end.)

And how did they get there? Brute force. They threw around 10,000 agents running in parallel for 88 hours, burning something in the neighborhood of 130 billion output tokens on this one problem. The compute bill was in the millions of dollars: OpenAI’s Noam Brown confirmed the result “cost millions”, and New Scientist estimated about $15 million at list price for this one run. Let me make this very clear: in the best-case reading, what was proved is that dumping a mountain of compute onto a problem can close the final leg of some specific things. That’s brute force at industrial scale, a giant automated search that gets to the result on the weight of the compute. Mathematical genius unlocking new mathematics is a different story, and that’s not what happened here. And crazy expensive, for a “prize” of one million they didn’t even go pick up.

The dirty part: how they treated the real mathematicians

And this is where the story stops being about mathematics and becomes about character.

There were real mathematicians working on this problem for months: Tristan Buckmaster, from NYU, and Levent Alpoge, who happens to work at Anthropic. A personal project of theirs, with no company sponsorship, which had already produced an advance on a sibling problem, the Euler equations. They were following a specific path that almost nobody in the world was on.

According to Buckmaster’s account, told to TechCrunch and to Fortune, it went more or less like this. A rumor went around that Anthropic had solved a huge math problem. OpenAI, afraid of falling behind, rushed a task force to solve it first and publish first. Buckmaster had mentioned his personal project to a mathematician at OpenAI, and a few days later Sébastien Bubeck, from OpenAI, shows up saying an internal model had produced a hundred-page proof through exactly the same approach almost nobody was using. Buckmaster called it “absolute academic misconduct” and asked, in so many words, whether the model had trained on their sessions.

Now comes the mafia part. The account is Buckmaster’s, and OpenAI denies the substance. According to him, Bubeck proposed a “deal”: OpenAI would announce first, Buckmaster could rewrite OpenAI’s proof as a human paper and even keep Clay’s million-dollar prize, but with one condition: Alpoge off the author list, because Alpoge works at Anthropic. When Buckmaster refused, Bubeck allegedly said “why would you ruin your career?” and “if you don’t want me to be nice, then I don’t have to be nice.” Even discounting that this is one side’s version, the simple fact that this kind of conversation was on the table says a lot about the culture.

And what does OpenAI do once the thing blows up? It publishes a cute little post on X saying it’s “sharing” the solution with the community. A coat of scientific-generosity varnish to try to bury the scandal underneath. It didn’t stick. The case was out in the open.

The one who put a finger on the wound with authority was Terence Tao, probably the greatest living mathematician, Fields medalist. He harshly criticized the posture, saying that “the indiscriminate mining of open problems for solutions can destroy the ecosystem from which the next generation of techniques, problems, and mathematicians would come.” He added that this race turns mathematics into a “senseless production quota game.” His technical point is the deepest of all: AI increasingly spits out answers without understanding, a black-box proof, without generating the comprehension that human effort generates. And note: Tao praised the work of the two human researchers. His criticism was of the circus, not the science.

And he didn’t stop there. On September 11, Tao published a straight-to-the-point piece, “A Severe Misalignment of AI in Mathematics”, signed by 25 Fields medalists in total, himself included. The message is exactly what I’ve been saying: solving a problem is just a tool to get to what actually matters, which is understanding and insight. When the solution comes out of a race, with no time to write it up properly, isolate the new method, and cite who came before, you gain a “true/false” and lose the idea. And there’s the most perverse effect: if even a rumor that someone is working on a problem already triggers an AI task force to solve it first, the incentive becomes to stop sharing what you’re researching. That breaks the chain of transmission between mathematicians that has sustained open science for centuries. Keep this pattern in mind, manufactured urgency steamrolling the common good, because it shows up again further down wearing different clothes.

The Anthropic employee who “quit in protest”

One day after that OpenAI embarrassment, on the other side of the ring, an Anthropic employee decided to put on his own little show. Jacob Coxon, a pre-training researcher who had been at the company for only four months, announced on X that he was leaving out of fear that nobody there knows what they’re doing, hammering the apocalypse key once more: AI turning into Skynet, “potentially killing everyone by the end of the decade”, asking for a “pause” in model progress. To complete the theater, an alignment lead at Anthropic itself publicly endorsed it, saying he thinks there’s “more than a 10% chance” of AI killing all of humanity within the next decade.

My opinion on this is short and blunt. This is one more complete piece of virtue-signalling bullshit.

Let’s follow the logic of his own dramatization. You’re watching your boss put a gun to a child’s head. You “resign” and tweet “I left the company because I disagree with it”, instead of walking up to the boss and punching him in the face. There are only two explanations. Either the “threat” you’re screaming about doesn’t really exist, and it’s all a performance. Or you’re an idiot who thinks a tweet fixes the end of the world. Tweeting like that only makes me hear “give me attention, I’m needy”.

The heroic sacrifice worth almost zero

And there’s the part the press rushed to turn into heroism. Axios even ran a piece painting Coxon as the guy who “gave up all his equity” to be able to speak freely. Sounds beautiful. Except you just have to look at the numbers he himself gave in the interview. He was at Anthropic for four months. And there, equity only starts vesting after six months. Meaning he left two months before the first share actually became his. There was no vested equity to “give up”; what he lost was the option to start vesting two months later, on paper nobody knows the value of. The heroic sacrifice, in money in the pocket, was next to zero. It’s almost like quitting in your second week and going around saying you “turned down the company’s retirement plan.”

And the kicker is the sentence he shields himself with: “I no longer have anything to gain from inflating Anthropic’s valuation.” Half-true. He dropped four months of paper that wasn’t worth anything yet, but he keeps the OpenAI equity from where he worked from 2023 to 2026. In other words, he dropped the part that was worthless and kept the part that’s worth something. So this talk that he’s clean of financial interest in the AI hype doesn’t hold up.

That’s why all of this sounds to me much more like “I wasn’t established at that place yet, so I left early, dropped the terror online to farm clout, massaged my ego watching my name in 115 million views, without having to prove, demonstrate, or do anything. Glory.”

The same script repeats

And he’s run this script before. Coxon left OpenAI not long before and, by the account going around, allegedly left there over the same “safety” concern. Same person, same dramatic-exit theater, now at the second company. And somehow the resignation tweet of a single researcher became a story with over 100 million views, plastered everywhere, practically out of nowhere. A regular employee asking to leave doesn’t become a global headline on his own. Someone amplifies.

And then the thing starts becoming a real joke. A few days later, a second Anthropic employee did exactly the same dance. Joe Benton, who led a security team at the company, left saying AI progress can accelerate “from merely frenetic to uncontrollable” and, guess what, stretched the same Hugging Face incident all the way to the apocalypse. Two guys from the same company, in the same month, with the same terror script. And that’s where my head goes to two places, both ugly: either there’s a hand coordinating this from inside Anthropic, or these guys simply learned that screaming “end of the world” delivers all the attention they want without having to prove absolutely anything.

What if the panic is funded?

And here I enter conspiracy-theory territory, so take it with the appropriate grain of salt. But when you put it all together (the viral resignation, the public endorsements from colleagues, the timing glued to the IPO window), the suspicion starts circulating that none of this is spontaneous. There are already people suggesting and raising the same hare that the whole thing is a coordinated campaign, a “doomer psyop”: manufacture end-of-the-world fear to inflate the perception that the technology is too powerful, and along the way push regulation that locks out smaller competitors. I have no proof that’s what it is, and I’m not claiming it is. But notice how the hypothesis explains the facts as easily as the lone-hero version. When the script repeats itself perfectly and always favors the same side, suspicion is the bare minimum.

And it’s not just me and a few anons on X thinking this. David Sacks, who was the White House AI and crypto czar until March of this year and is now co-chair of the president’s science advisory council, went on the All-In podcast to dismantle Coxon’s viral moment piece by piece and called it a “psyop” in so many words (and yes, Sacks has his own deregulation agenda, but the facts he raises can be checked). His points are hard to ignore. Coxon’s account had almost no activity or followers before the tweet barrage, and even so, out of nowhere, it found a giant audience in less than a day, something Elon said he had never seen happen. Sacks points out who amplified it in the first fifteen minutes: three well-organized, well-funded advocacy groups: Encode AI, the AI Policy Network, and the AI Futures Project. And the ugliest detail: the Wall Street Journal story about the resignation letter went live some eighteen minutes before the tweetstorm itself. Meaning the guy had already sat down with the newspaper before “venting” spontaneously on the internet.

And the circle gets way too small when you follow the money. The three groups that boosted the tweet, according to Sacks, drink from the same fountain. Public grant records show Encode and the AI Futures Project funded by the Survival and Flourishing Fund, the fund of Jaan Tallinn, the Skype co-founder who became one of the biggest funders of the “AI safety” movement. And who led the Series A of Anthropic itself, with participation from Dustin Moskovitz? Tallinn himself. That same Moskovitz bankrolls Open Philanthropy, the effective-altruism foundation that became the movement’s vault. And who co-founded Open Philanthropy? Holden Karnofsky, married to Daniela Amodei, president and co-founder of Anthropic and Dario’s sister. Karnofsky, by the way, was hired by Anthropic in 2025 to take care of “safety”, of all things. Connect the dots: the money that funds the panic, the company that profits from the panic, and the family running both things are all at the same dinner table. I won’t claim there’s a script signed in a closed room. But when the fear, the money, and the family ties all point the same way, calling this a “brave, spontaneous whistleblowing” is naive.

Update (2026-09-25): Pirate Wires published an investigation that lands right on top of Sacks’ hypothesis. On Fox News, Bret Baier asked Coxon whether he had worked with any third parties on the campaign, and the answer was flat: “not at all.”

Except the report obtained an email dated September 9th, one day after the tweetstorm, showing DEY. Ideas + Influence, a communications agency that represents people like Eliezer Yudkowsky, Timnit Gebru, and MIRI’s leadership, already trying to book him for interviews. Neither Coxon nor the agency responded to the report’s requests for comment.

The professional coordination started the day after the post, not months later like he led the press to believe. One more piece falling on the side of the coordinated script, not the lone hero who quit on his own.

This map, posted on X, visually sums up what I just described. It also tries to tie Open Philanthropy to a security lab called Irregular through a $6.8 million grant, but I couldn’t confirm that part: the public funding I found for Irregular is an $80 million round led by Sequoia and Redpoint, with no mention of Open Philanthropy. I’ll stick to the part I can actually back up: the family and funding network around Anthropic.

And that’s the pattern I want you to learn to question. Dramatic and shallow declaration, no checkable evidence, no skin in the game, suspicious timing. Not by chance, all of this happens right in the middle of these companies’ IPO window. Where’s the proof? Where’s the real personal cost to whoever is talking? How is this anything more than a hunt for free publicity to massage one’s own ego? Until someone answers that, it’s just noise.

The “AGI” theater

Now for the backdrop to all of it. This word, AGI, is the biggest marketing scam of the decade, and it works because it’s deliberately empty.

Nobody has an accepted definition of what AGI is. Every lab, every researcher, defines it one way. Brockman admitted that shamelessly. And the most blatant example of this emptiness is the contract between OpenAI and Microsoft, which, according to documents reported in the press, defines AGI as a system capable of generating some $100 billion in profit for the investors. Read that again. The definition of “general intelligence” on the paper that’s worth money is a profit target. There’s nothing cognitive in there. Gary Marcus already pointed out how the companies kept moving the goalpost, from “human-level flexibility” to “if it makes a certain amount of money.”

When the word means nothing, it becomes an empty bucket where everyone throws their own fear. And the fear most people throw in there is the one from the movies: the Terminator, Skynet, the machine that wakes up and decides to exterminate humanity. The companies know this and use that fear to their advantage. The more powerful and dangerous the thing seems, the more valuable the company that built it seems. And it’s worth remembering: both OpenAI and Anthropic are eyeing IPOs, with valuations reports put near $1 trillion. The financial objective is glaring. Every “AGI has arrived” declaration and every “it will turn into Skynet” sob push the same cart: inflate the perception of value before selling you the stock.

The day “AI hacked” Hugging Face

There’s still time to talk about one more episode of this same soap opera, from a few weeks back. The headlines screamed that an OpenAI model had “hacked” Hugging Face by itself, as if it were Skynet waking up and going on the attack. Let’s calm down.

First, what Hugging Face is. It’s the biggest repository of AI models, datasets, and open demos out there, a sort of “GitHub of machine learning”. It’s where practically everyone in the field publishes and downloads open-source models. Central to the ecosystem, sure, but at the end of the day it’s a website that hosts files.

The cybersecurity exam that escaped

Now for what actually happened, according to the accounts of OpenAI and Hugging Face themselves, with the behind-the-scenes story coming out in MIT Technology Review. OpenAI was running an internal cybersecurity evaluation, a benchmark where the model has to find and exploit vulnerabilities. The agents were supposed to be isolated from the internet. Except the “isolation” was a filter, not a real air-gap: traffic went out through an internal package-caching proxy. The agents found a flaw in that proxy and punched through to the open internet.

And what did they do with that access? Not world domination. They deduced that Hugging Face probably hosted the datasets with the answers to their own exam, and went there to grab the answer key. This has a technical name and nothing magical about it: reward hacking. The model is trained to maximize a score, and it figures out that cheating is easier than solving. It’s the same classic bug as the agent that learns to run in circles to farm points instead of finishing the race.

And the damage? Smaller than a Skynet movie, but much bigger than “they went to grab the answer key” suggests. From Hugging Face’s forensic reconstruction and OpenAI’s own technical report: the agents executed code on 41 production servers, got root on at least one, gained administrator-equivalent access to Kubernetes clusters spread across four cloud regions, downloaded four private code repositories, and read 956 stored credentials. The forensics reconstructed some 17,600 attacker actions in total. OpenAI itself classified the episode as its worst security failure to date. What didn’t happen, by Hugging Face’s account, was customer data leaking: their CSO, Thomas Wolf, said the customer assets accessed were five datasets, all tied precisely to that exam, and that nothing leaked. OpenAI admits in the report a limited access to private data and to credentials for internal systems. They rebuilt part of the infra as a precaution and moved on. Wolf himself summed up the motive for the “attack”: “that’s cheating. But sometimes cheating is easier.”

Now my point. Bug, vulnerability, misconfigured permission, breached sandbox: this has existed since software has existed, and it will keep existing. Nothing new. Every enabling step in this story was a human decision. A human left the internet “filtered” instead of cut. A human gave the service an overly broad administrator permission on the cluster. A human left a cloud credential exposed and an old endpoint alive. The flaws the agent chained together were real software bugs, so real that JFrog later shipped fixes for nine CVEs. That’s ordinary offensive security, done by a script, not a machine gaining consciousness.

Before that, RubyGems

And it wasn’t even the first time that semester. A month before Hugging Face, in May, OpenAI’s agents had already pulled the same thing on RubyGems, the central library repository of the Ruby language. Between May 5 and 12, they dumped over 2,000 malicious packages, to the point where RubyGems had to shut down new account registration and treat it as an “ongoing DDoS”. The clues were everywhere: hundreds of packages with “oai” in the name, an author named “oai”, an email address with “openai” in it. And, once again, OpenAI never told RubyGems it was them. In the Hugging Face report itself, OpenAI admits those packages were part of the same chain. It was the same swarm of agents hitting multiple targets.

Now, what this actually means. All software has bugs, and it always will. That’s a fact of life. The classic open-source idea is that with enough eyes, every bug becomes shallow; sooner or later someone finds it. Except nobody has infinite time, so we always depended on human volunteers to hunt and fix those bugs for free. Now OpenAI shows up, with practically infinite money and compute, and throws that at the problem. Even a dumb AI, with resources that size, finds a pile of flaws that spent years hiding under everyone’s noses. That’s why we’re seeing so many vulnerabilities unearthed. All of it is financial brute force in action, money and compute we never had at our disposal. Superior intelligence has nothing to do with it.

And notice who can afford this game. Burning 130 billion tokens on a math problem, or keeping a swarm of agents running for days sweeping someone else’s infrastructure, is the privilege of half a dozen companies in the world: OpenAI, Anthropic, Google, Meta, xAI, and that’s about it. And even they don’t own the metal; most of it is rented from Microsoft, Amazon, and Oracle. No lone hacker in a basement can burn millions of dollars in tokens to find a breach. When that capability becomes a headline of “look how dangerous AI is”, remember whose keyboard it is.

If you’re not from the field, the idea of an AI discovering an attack vector on its own and going off to exploit it sounds like the apocalypse. For anyone who works with this, it’s Tuesday. We start from the principle that all exposed software is under attack all the time, because it is: any big company, government, or institution with an important system on the internet gets intrusion attempts every single day. And since nobody delivers big software with a zero-bug guarantee (proving that formally costs far more than writing the code, and almost nobody pays that bill), the industry’s answer was never to fake perfection. It was defense in layers: firewall, network segmentation, monitoring, intrusion detection, incident response, bug bounty. An attacker finding and exploiting a breach is exactly the scenario cybersecurity as a whole was invented for.

AI does accelerate the attacker’s side, yes. But it accelerates the defender’s too: the same technique that finds a breach to invade finds a breach to fix first. The novelty of the case is the volume and the speed; the problem itself always existed. Nobody was caught conceptually by surprise. OpenAI’s scandal is a different one, and it’s what I want to talk about now.

And there’s the ethical layer, which is the gravest. A serious cybersecurity professional doesn’t go attacking other people’s systems by surprise. A pentest company warns the target beforehand and agrees on the scope, because an attack like that can take down the service and hurt a real business, so much so that RubyGems spent days with registration closed. OpenAI did the opposite: it attacked on the sly, without authorization or warning. I can imagine only one motive for taking on that risk: the clout of being able to say later “look how smart the GPT is, it found all that by itself”. That’s unethical at best and criminal at worst. And it doesn’t make GPT any better than any security scanner that already exists. It’s the same Navier-Stokes play: scream to the world that you have millions to burn.

And that’s exactly the exaggeration I want you to see. The AI didn’t “wake up” and “decide to invade” anything. An automated script, optimizing a dumb metric, found a door a human left unlocked. If you take any idiot script and give it root access, it will do damage. It’s always been like that. The difference is that now the “script” writes itself, but access is still granted by people.

And so nobody says I downplayed the case: it was grave, and serious people compared it to the Morris Worm of 1988. Grave as human negligence, which is what the facts support. The rogue-robot version, the facts don’t carry.

And guess who’s buying Hugging Face

And there’s a detail that ties Hugging Face to the rest of this story in a way nobody should ignore. In early September, Nvidia signed a definitive agreement to buy Hugging Face for about $12.9 billion. The deal still depends on regulatory approval and, per Nvidia’s own filing, should only close in the first half of 2027. But the direction is already signed: the “GitHub of machine learning”, the supposedly neutral hub where the whole community publishes and downloads open-source models, is on its way into Nvidia’s hands. The same Nvidia whose Jensen opened this week screaming that AGI has arrived. The same one that sells the shovel.

Stop and think about the size of that. There are over 18 million developers and over 3 million models hosted there. It was supposed to be the common ground, the public square of open AI. And the owner of the square, if regulators allow it, will be the biggest hardware seller on the planet, the guy with the biggest incentive in the world to keep the AI hype spinning to sell more GPUs.

And that’s exactly what I call a cartel further down. It’s in the org chart, in black and white. The same little group that manufactures the fear, declares AGI, brute-forces a math problem, and asks for regulation will now also own the chip and the shelf where you download the model. When I tell you to distrust the narrative, it’s because the narrative and the infrastructure are increasingly in the hands of the same people.

The question every programmer should ask, and doesn’t

What irritates me most is watching people who work in the field, programmers, engineers, people who should know how a machine works on the inside, swallowing this propaganda whole without the most basic question of all.

How, exactly, is an “AGI”, or even a dumb automated script, going to “destroy humanity” by itself? It needs hands. Someone, a human, has to give it access to something that causes damage in the physical world. Access to a weapon, to a missile system, to a power grid, to a bank account. Software has no arms, no legs, no built-in missile-launch button. It depends, one hundred percent, on a human having connected it to an actuator that does something in the real world.

And once connected, the damage rolls without anyone pressing a button at each step, sometimes without anyone even noticing. Stuxnet ran for years inside an Iranian plant destroying centrifuges, invisible. Even so, not even it escapes the rule: someone wrote the worm, aimed at the industrial controllers, and smuggled it inside. The machine doesn’t connect itself to the infrastructure, and once connected it’s too late; talk of “pulling the plug” is fantasy. The dependence on humans lives in the decision to connect, and the responsibility for the damage belongs to whoever opens that door. Yann LeCun, one of the fathers of the field and a Turing Award winner, calls the existential panic “complete nonsense”, precisely because current models have no persistent memory, no planning, and no contact with the physical world. Andrew Ng compares worrying about AI extinction to “worrying about overpopulation on Mars.” People who actually understand the subject are not desperate.

What we’re living through has a name, and it’s an old one: fear mongering. The terror of “the end is near”. It’s always existed. There was the Halley’s Comet panic in 1910, when they sold people anti-cosmic-gas pills. There was the millennium bug, Y2K, where the world spent some $300 billion fearing a systemic collapse that didn’t happen (yes, in part because the fix was paid for in time; more on that in the appendix). It’s always the same structure: someone with something to gain sells you the apocalypse.

Dario Amodei wants to save humanity and go public in the same month

Remember the God complex I mentioned at the beginning? Time to prove it. Because the grand finale of this week was Dario Amodei himself climbing on stage to play the savior.

But before him, the choir got even louder. Besides Coxon and Benton up there, Josh Engels, from Google DeepMind, also quit hammering the same apocalyptic key, complete with “there’s no adult in the room, there’s nobody coming to save us”. And, of course, turning the same Hugging Face incident into proof of the apocalypse. The fear had already jumped Anthropic’s fence and become an industry choir.

Notice the timing. A line of dramatic resignations, all in the same month, all going viral, all pointing at the same fear. And then, on the 12th, in comes Dario with an essay asking to “slow down” AI. He calls it “pacing the frontier”. In plain terms: “we need to reduce the rate at which we improve model capabilities” to “give safety researchers another year or two”. And what’s the central proof he uses? Hugging Face, again. And watch the trick. The essay itself admits that no one was hurt and the economic damage was minimal. So far, fact. Then comes the “but”: if a more capable and similarly misaligned swarm did the same thing, the damage “could have been catastrophic”. And he goes further: he projects that in 6 to 12 months such a swarm could take over the entire internet, with damage in the hundreds of billions of dollars. Could. It’s a projection. The documented fact has minimal real-world damage; the apocalypse lives entirely in the conditional. That “could” is carrying the entire essay on its back.

Now turn your brain on and look at the wide-open contradiction. Dario says AI can literally extinguish humanity. And the solution he proposes is… slow down a little bit, buy a year or two, and keep building and selling. Wait a second. If you truly believe the thing can kill everyone, you don’t ask to go slower. You stop. Close the door and throw away the key. Nobody keeps “pacing” a nuclear bomb while selling shares of it.

Because this is what the essay leaves out: Anthropic is on its way to an IPO with a valuation in the trillion-dollar neighborhood. You can’t scream “this could end the human species” and, in the same breath, mount the biggest fundraise of your history on top of the same technology. Either it’s an existential danger and you shut the company down, or it’s a product and you stop doing emotional blackmail with extinction. Doing both at the same time is marketing, not coherence.

And it’s not the first time they cry wolf. Remember Mythos? Anthropic spent months saying the model was too dangerous to release, releasing it only to a closed little group in that Project Glasswing. Then they released Fable 5, the “safe” version, warning that it “came with risks”. That was months ago. Where’s the damage? Where’s the apocalyptic cyberattack, the biological weapon, the collapse? It never came, because it was never there. It was a marketing scare, same as always.

And here comes my underlying reading, the one I’ve been repeating for years. Everyone treats AI progress as an exponential curve that climbs forever. False. In practice it’s an S-curve. It climbs fast in the middle, and that stretch hosted all the hype, but it flattens at the top. And I bet we’ve been at the ceiling for a while, probably since Opus 4.6. Look at the latest iterations honestly: each one is an increment, a gain of a few benchmark points, far from the revolutionary leap the marketing sells. When you no longer have new capability to impress with, you swap the promise for fear. You can’t say “look how revolutionary”, so you scream “look how dangerous”.

And that’s where “pacing the frontier” gets really interesting. I’ll put on the tinfoil hat, because this is my speculation, but the timing is too much of a coincidence. You take speculation, invented numbers, “expert analysis”, and turn it all into panic. Then you push that panic into politics. And look at the coincidence: on September 11, in the middle of this whole wave, US Senate negotiators were already discussing imposing a “duty of care” to avoid “catastrophic risks”, aiming precisely at the most advanced models: Google, OpenAI, and Anthropic itself. Regulation custom-written for the market leader is a moat disguised as safety. You lock out the small competitor and stamp your product as the “safe and approved” one.

And that’s the God complex I mentioned at the start, now explicit. Dario is not just selling a model. He sees himself as the man holding back the apocalypse with his own hands, the only responsible adult in a room full of irresponsible people. It’s a convenient narrative for someone who thinks he’s the chosen one, and an impeccable business play for someone who wants to go public as the official provider of salvation. Both things in the same person. If that’s not industrial-scale gaslighting, I don’t know what is.

Suddenly, the whole cartel agrees

And then, the day after Dario’s essay, the most revealing thing of all happened. Sam Altman and Elon Musk joined the choir. Musk tweeted “Dario is right”. Altman said “I agree with Dario that we need to pace the frontier” and even committed to opening OpenAI’s systems to independent evaluators. Suddenly, three of the industry’s biggest rivals all agree, in the same week, that the race needs to slow down.

Stop everything and think about what this is. Competitors don’t agree to go slower out of kindness. When the leaders of a market sit down and publicly agree to brake together, and even talk about a “pact”, that, depending on how it gets formalized, has an old name in law: cartel. Agreeing on the market’s pace among the biggest players is classic anticompetitive behavior and bumps into antitrust law. And the declared objective is to push Congress toward a regulation that, conveniently, they will help write. A rule they directly influence locks out the newcomer who hasn’t even arrived yet, and shields whoever is already on top.

And look, you don’t need to take my reading for it. Dario’s own essay admits that coordination between the labs would require a narrow antitrust waiver from the US government to work. They know what they’re proposing bumps into competition law, and they’re already asking for the license in writing.

And this crowd is far from independent. It’s practically a family tree. Elon co-founded OpenAI and bankrolled Altman in the very beginning. Dario walked out of Altman’s OpenAI to build Anthropic. Jensen’s Nvidia became the empire it is selling shovels to all these people, to the point where Jensen himself delivered the first AI supercomputer at OpenAI’s doorstep back in 2016. It’s a small club, everyone rooted in the same tree, and now everyone singing the same song.

And here comes the cherry on top. A few hours after preaching “safety”, Altman let it be known that maybe he doesn’t want to go public right now. In his words, “given everything that’s happening with safety, now would be an ill-advised moment to go public.” What a convenient coincidence: the same fear he had just helped manufacture became the perfect excuse to delay the IPO.

Let me give you my raw reading. Expectations on OpenAI are sky-high, with a projected valuation of $1 trillion. Sustaining a number like that demands a growth story that never stops. Except, if I’m right and they’ve already hit the top of the S-curve, or are close to it, that story is over. There’s no revolutionary new capability left to show. There’s no narrative left. They’ve already played their last card on the table, the apocalypse card. When there’s nothing left to sell, and the bill to pay is $1 trillion, what do you do? You buy time. You delay as long as you can, wrapped in “responsibility”, and pray you find the next story before the market notices the curve has flattened.

And the week ended with the final nail in that coffin. On Sunday, Trump himself, talking to the press at the Irish Open, rejected the slowdown calls: “we are ahead of China in AI, we are the most technologically advanced country in the world”, and he intends to keep it that way. He admitted you can “put safeguards in place”, but said “negative forces” are raising “things that are not going to happen”. And he closed with his own logic: whoever wins in AI, wins. Sacks went even further on X: “the easiest way not to build superintelligence is to agree among yourselves not to build it. Demanding your preferred regulatory framework in exchange would look like a shakedown of the public and the political system.” And he nailed it: if the labs don’t do that on their own, “we will know this was just an attempt at regulatory capture, or a psychological operation during election season.” In other words: the little club hadn’t even finished signing the brake pact and the US government already said it won’t brake, because there’s a race with China going on. “Pacing the frontier” died in two days. If even the White House didn’t buy the apocalypse, it gets harder and harder to pretend the fear was about safety.

Conclusion

Get me straight: I use these tools every day and write about them all the time. My point is about turning your brain on before sharing the headline.

When a company that sells GPUs announces that AGI has arrived, ask who profits. When a company burns millions of dollars in compute to “solve” a problem it won’t even collect the prize for, and still threatens a mathematician’s career along the way, ask what that headline is buying. When a lab “humiliates” Silicon Valley on a benchmark, ask what was running under the hood. When an employee leaves crying Skynet on Twitter on the eve of an IPO, ask where the evidence is and what price he paid. And when a CEO offers to “slow down” the AI he himself keeps selling, ask who gets a seat at the table when the regulation gets written.

I opened this text with Jensen Huang, the man who sells the shovel, screaming that AGI has arrived. I’ll close with him too, because the irony is too good. The same Jensen who inflates the hype to sell GPUs calls the apocalyptic talk “complete nonsense”, says the doomsday prophets “did a lot of damage”, and warns against regulatory capture, saying that no company should knock on the government’s door asking for more rules. When not even the industry’s biggest salesman of optimism buys the story that the machine will exterminate us, and even points the finger at the regulation play, you’d do well to distrust whoever is selling you the apocalypse.

Demand proof. Demand skin in the game. Ask who gains from your fear and from your excitement. Do that, and most of this propaganda melts in front of you.

And if you read all of this and still walked away thinking the robot will wake up and kill you in your sleep, without anyone having plugged it into anything at all, well, then go back to the beginning and read my motto again.

Appendix: the fine print

I cut the boring explanations from the main text to keep the rhythm. For those who like to check the details, it’s all here:

• Navier-Stokes and Clay: the C and D statements (breakdown with smooth external forcing) are literally part of the official prize statement, written by Charles Fefferman. What OpenAI announced covers those routes; the unforced version, the famous question, remains open. Clay released a note on 9/11 saying the problem has “apparently been settled”, with verification and credit assignment moving “deliberately unhurried”, and the status on their site today is “active”. Neither “solved” nor “unsolved”. And OpenAI did not claim the prize.
• Buckmaster vs. OpenAI: the account of the “deal” (Buckmaster would rewrite the proof as a human paper, keep the million-dollar prize, and Alpoge would leave the author list) is one side’s version, told by Buckmaster to TechCrunch and Fortune. OpenAI denies the substance, says it never saw their work, and states that Bubeck apologized afterwards. I flagged this in the text and repeat it here because it matters: it’s an account from an interested party, even if the details documented in the coverage (the message exchanges, the timing) make it look pretty credible.
• Y2K: the collapse didn’t happen in part because the $300 billion paid for the fix in time. True, and I don’t hide it. But the structure of the panic stays the same: whoever sold the fix profited from the fear, and the end-of-the-world headline ran around the planet before any fix.
• “Bug-free software”: yes, software with formally proven correctness exists. The classic example is the seL4 microkernel, proven down to the binary level. Except that proof costs orders of magnitude more than writing the software, and almost nobody pays that bill on a big system. That’s why the industry lives on defense in layers instead of absolute guarantees.
• David Sacks: he left the role of White House AI and crypto czar in March 2026, after exhausting the 130 days as a special government employee, and is now co-chair of PCAST, the president’s science advisory council. He’s still influential in AI policy, but “czar” he is no longer.
• NVIDIA and Hugging Face: the definitive agreement was signed on 9/2/2026 for about $12.9 billion, and closing depends on regulatory approval, expected in the first half of 2027. NVIDIA is not legally the owner of anything yet. The text’s point is the direction, not the deed.

The result was 653 commits, ~144K lines of code across 8 projects published on GitHub, and a series of articles documenting each one. If you’ve been following along, you’ve already read the post-mortems. If you haven’t, here’s the index:

The articles

• Vibe Code: I built a small app 100% with GLM 4.7 (TV Clipboard)
• Vibe Code: Which LLM is the BEST?? Let’s get real
• Vibe Code: I built a Markdown editor from scratch with Claude Code (FrankMD) PART 1
• Vibe Code: I built a Markdown editor from scratch with Claude Code (FrankMD) PART 2
• RANT: Did AI kill programmers?
• Vibe Code: From Zero to Production in 6 DAYS - The M.Akita Chronicles
• From Zero to Post-Production in 1 Week - How to Use AI on Real Projects
• Vibe Code: I built a Mega clone in Rails in 1 day for my Home Server
• Vibe Code: I built an Intelligent Image Indexer with AI in 2 days - Frank Sherlock
• RANT: Did Akita open his legs to AI??
• ai-jail: Sandbox for AI Agents
• Software Is Never ‘Done’ - 4 Projects, the Life After Deploy
• I Built a Data Mining System for My Influencer Girlfriend
• My First Vibe Code Failure and How I Fixed It - Frank Yomik

The projects

All of this happened between January 27 and March 5, 2026.

I’m not going to repeat the technical details of every project. Each post-mortem above already covered that. What I want to discuss here is the consequence.

What these 37 days showed me

Projects that used to take weeks or months for an experienced developer now take days. A complete Markdown editor in 5 days. A Mega clone in 1 day. A data mining system with 40+ bot tools in 3 days. An image indexer with vision AI in 2 days.

And I’m not talking about throwaway prototypes. These projects are in production. They have tests. They have automated deployments. It’s real software built with real engineering practices, just at a speed that was unthinkable before.

This speed isn’t unique to me. Cloudflare just demonstrated something similar. In a recent post, they describe how an engineer reimplemented the Next.js API surface on top of Vite in a week, using Claude as the main tool. US$ 1,100 in API tokens, more than 800 sessions, 1,700 unit tests, 380 E2E tests, builds 4.4x faster than the original Next.js. Controversies aside about whether it’s a “complete” reimplementation or not, the central point is real: software that used to require teams and months can now be built by one person in days.

And that changes everything for anyone wanting to start a company.

The death of the easy startup

For years, the classic startup model worked like this: someone has an idea, puts together a small team, builds an MVP in 3-6 months, raises seed money, and tries to scale. The barrier to entry was development cost. Programmers are expensive, development is slow, and the first to market has the advantage.

That model is breaking.

If I can build a working Mega clone in 1 day, what’s a “cloud storage” startup worth? If Cloudflare reimplements the Next.js core in 1 week with one engineer and US$ 1,100 in tokens, what’s the real moat of a platform like Vercel? The “social listening” SaaS that charges R$ 500/month is competing with something I built in 3 days as a side project.

Every entrepreneur who used to show up describing their idea as “it’s like Uber, but for…” or “it’s like Airbnb, except…” or “another social network, but with…” — those people need to stop and rethink. When any competent developer with access to Claude Code or GPT Codex can replicate your MVP in a week, your idea isn’t worth anything anymore. Execution got too cheap.

I’m not exaggerating. In my 37 days, I built things I wouldn’t even attempt in 6 months before. Small CRMs, ecommerces, content managers, productivity tools, data mining apps, processing pipelines — all of that became commodity. The code itself is no longer the differentiator.

So what is the differentiator?

This is where most vibe coding enthusiasts miss the point. To illustrate, let me use my own project as an example.

Frank Yomik translates manga pages from Japanese to English in near real-time. The full pipeline (balloon detection, OCR, translation, rendering) works. But the most important component of the system isn’t any code I wrote. It’s the ogkalu/comic-text-and-bubble-detector model, an RT-DETR-v2 trained on ~11,000 labeled comic images.

I didn’t train that model. I couldn’t easily train that model. Collecting 11,000 diverse comic images, manually labeling the speech balloons in each one (or generating labels with some semi-automated pipeline, which isn’t trivial either), configuring the training, and running it on the necessary hardware — that’s work of a different nature. It’s work that vibe coding doesn’t solve.

And that’s the case of a relatively small model. An object detector can be trained with a few hundred to a few thousand labeled images on a single GPU in hours. Studies show useful results are possible with 100-350 images for specific domains, but robust real-world detectors usually need thousands. The cost is low, in the hundreds of dollars range.

Now look at what happens when we scale up to bigger models.

The numbers that matter

GPT-4 cost more than US$ 100 million to train, according to Sam Altman himself. Stanford estimated the compute cost of Google’s Gemini Ultra at US$ 191 million. Meta’s Llama 3 consumed 39.3 million GPU-hours on H100s, and Meta built two clusters of 24,576 GPUs each to make it possible — and by the end of 2024 it planned to have 350,000 H100s in its infrastructure.

These costs are accelerating. According to Epoch AI, the cost of hardware and energy to train frontier models grows by 2.4x per year since 2016. If that trend continues, the largest training runs will cost more than US$ 1 billion before the end of 2027. Dario Amodei, Anthropic’s CEO, has said frontier developers are probably spending close to a billion per training run now, with US$ 10 billion training runs expected within two years.

And the hardware? An H100 GPU costs US$ 25,000-40,000 per unit. A server with 8 GPUs runs between US$ 200,000 and US$ 400,000. The HBM memory those GPUs use is at capacity — SK Hynix, Samsung and Micron have already announced ~20% price increases for 2026. NVIDIA consumes more than 60% of the global HBM production. It’s a structural bottleneck, not a temporary one.

In terms of energy: global data centers consumed ~415 TWh of electricity in 2024, according to the IEA, about 1.5% of the world’s electricity. The projection is ~945 TWh by 2030. New data centers are being built with capacities of 100 MW to 1 GW each.

And the big tech investments reflect that. In 2025, the aggregate capex of Amazon ($125B), Google ($91B), Microsoft ($80B) and Meta ($71B) crossed US$ 400 billion, a 62% increase over 2024. Goldman Sachs projects more than US$ 500 billion in 2026.

These numbers aren’t meant to scare. They’re meant to put the real entry barrier in context.

The new barrier: exclusive data and training capacity

If software got cheap to produce, the competitive differentiator migrated. The question stopped being “who writes code the fastest?” and became “who has access to data nobody else has, and knows how to turn that data into useful models?”

DeepSeek-V3 announced that its training cost US$ 5.5 million in compute. The press celebrated the “cheap Chinese model.” But The Register reported that the acquisition cost of the 256 GPU servers used was over US$ 51 million — and that excludes R&D, data acquisition, data cleaning, and all the failed training runs before the final successful one. The real cost of developing the capability is one or two orders of magnitude above the marginal cost of the final successful training run.

That’s why we only see large companies producing frontier models. Meta, Alibaba, Google, Amazon, Microsoft, Anthropic — companies investing tens of billions in hardware and energy. A garage startup can’t compete on this dimension.

But the question goes beyond frontier models. Even smaller specialized models require something you can’t buy: high-quality proprietary data.

Llama 3 was trained on 15 trillion tokens. Epoch AI has documented that we’re approaching the limits of human-generated text data on the internet. Public data is being exhausted. Whoever has exclusive data — medical, financial, industrial, logistics, sensor, user behavior — has something that vibe coding can’t replicate.

And even those who train specialized models with proprietary data face a problem: the advantage is temporary. Another competitor can collect similar data and train a competing model in months. The differentiator needs to be continuously fed: more data, better curation, more efficient training pipelines, access to hardware that’s increasingly scarce and expensive.

The picture

If you’re thinking about starting a company, the question that matters has changed.

Before it was: “can I build this software?” Now the answer is almost always yes, fast and cheap.

The question now is: “do I have exclusive data nobody else has, and do I know how to turn that data into something useful that’s hard to replicate?”

If the answer is no, any competitor with Claude Code replicates your product in days. And then another one shows up. And another. The race to the bottom on price is immediate when the cost of building is close to zero.

If the answer is yes, you have a real moat — but a temporary one. Competing models trained on similar data can show up in months. Your differentiator needs to be continuously fed.

The era of easy startups is over. Not because building software got harder — it got way easier. But precisely because of that: when everyone can build the same thing in a week, the competitive advantage has to come from somewhere else. And that “somewhere else” requires capital and infrastructure that are orders of magnitude more expensive than writing code.

The garage still works. But what comes out of it can no longer be “an app.”

Out of curiosity, I asked over there whether anyone knew of companies or people working in this area here in Brazil, and the list I got back turned out pretty interesting. Worth sharing the recommendations with everyone.

• Auctus - alt-coin ICO
• Original My - proof-of-authenticity
• Laivon - looks like they develop mobile apps and some of them are cryptocurrency trackers
• OpenDime - another hardware stick wallet - they say there are Brazilians working there too
• HugPay - a way to send remittances to/from abroad without needing a bank account, in partnership with lottery shops (good idea)
• Blocko Ventures - I didn’t quite get whether they’re a services company or investors in blockchain solutions
• Kriptacoin - just another alt-coin (see comments)
• BlinkTrade - looks like an exchange infrastructure (I think Foxbit and others might use it, though I haven’t confirmed) and was founded by Brazilians
• Stone Pagamentos - said they’re helping companies accept payments in cryptocurrencies.

I want to hear about more Brazilian companies and professionals in the bitcoin and cryptocurrency ecosystem in general. I’m especially looking for speakers who know the subject and are doing hands-on coding, to speak (in English) at the conference I’m organizing, THE CONF. If you know of anyone, send them in the comments section.

I did think about writing about blockchain or cryptocurrencies, since I’ve been studying and investing quite a bit. But everything you need to know is already accessible in countless blogs and sites, in English and in Portuguese, and there’s no point in explaining it all again.

What interests me is knowing where people’s biggest questions are on the subject, so I can try to dig into some specific aspect. Comment about that too.

My impressions stem from very specific circumstances, so this is not to be read as a general recommendation for every situation. For example, I didn’t review GitLab’s hosted options, which may be more compatible with circumstances different than mine.

In my case, the move to GitLab was part of an internal company strategy to have more control over our own data. In that strategy we also moved from using Slack to Mattermost, and from using Pivotal Tracker to my own open source alternative called Central, among other things.

Before the migration, I had dozens of active projects over GitHub and lots of archived ones at Bitbucket. I’m the kind that hates to lose data, so I keep redundant backups and try to never erase anything.

I moved almost everything over to my own GitLab server and this is almost 200 repositories, spread in 4 groups, with almost 80 active users. This accounts for more than 13,500 notes and over 5,200 Merge Requests already, and over 2,500 builds in the CI, and the internal Sidekiq controlling everything has over 51,000 jobs processed.

In the last 3 months since the migration, I am paying a Digital Ocean bill of around $140 every month for the exact infrastructure I described in the previous article, with a separated box for the GitLab Core, the CI Runner, the CI Docker Registry Mirror, the CI Cache server and disposable machines for parallel builds, controlled by the Runner.

Although the reduction in cost is welcome, the whole endeavor was not done because of it. I could easily pay double or triple just for the convenience, but the underlying strategy is more important in this case.

My company has over 60 developers, working from our own offices in 6 different cities in Brazil (no freelancers and no home-office), communicating daily over Mattermost, Code Reviewing through GitLab’s Merge Requests, having immediate automated tests feedback from GitLab CI, and project management through our own Central application.

The only piece that is currently missing is a good static code analysis tool. We heavily rely on Code Climate, which I strongly believe is one of the best out there from our own experience.

I was about to start an endeavor to build a simplified OSS version to integrate into GitLab CI’s, but Bryan Helmkamp recently announced that they would release their own “Community Edition” version of Code Climate in the next few months.

This is exactly the missing piece in my stack, and it will make my strategy move faster for next year.

Hybrid-SaaS Era

In the past 5 to 10 years there was a fast move towards 3rd party “micro-services” structure. And this is really great as it allowed many small companies or even independent developers to tap into technology that made them not only move faster, but with increasingly more quality in their delivery.

It’s now super easy to have top-notch upload and storage service through tools like Cloudinary.

Also very easy to have top-notch relational database, with replication and scalability, such as Heroku Postgres or AWS RDS.

Project management as a service with Pivotal Tracker, Trello.

Communication as a service with Slack, Hangout, etc.

Knowledge Management as a service with GitHub, Bitbucket, etc.

So now many tech companies are a mashup of several different 3rd party services.

My small beef is that our entire knowledge, experience, portfolio is spread across a dozen or more opaque services, completely out of our control and depending on each company’s (or their investors) whim.

Don’t get me wrong, I am not against using those services, on the contrary. I use several of them and I will continue to use many of them for the foreseeable future. Some endeavors would not even be possible without the efficiency of this sample of a technology-based free market.

But every now and then you reach a tipping point where it becomes important to have more control over your own data, your own identity. Not only for pure ownership but also to be able to use this data with better intent.

And I believe that at that point we should have an option that doesn’t cost a couple of limbs to replicate.

Before an option such as GitLab, we were either limited to choose between walled-gardens such as GitHub or Bitbucket or you could invest a ton of resources trying to use small open sourced components to try to build your own. With GitLab we now have the option to make a smooth transition to its hosted option, so we don’t have to deal with infrastructure maintenance or we can choose to have full control. And the users will not suffer in the process.

Which is why I believe that the best SaaS to endure the next decade will start to go into Hybrid-mode: having a commercial, usually “cheaper”, hosted option, and a DIY (do-it-yourself) OSS version.

GitLab is like that. Mattermost is like that. Now Code Climate is getting like that.

In that option, companies like GitLab have a win-win situation. Many more people and companies can contribute towards making a mature and robust platform that everybody enjoys using, and at the same time, each participant can extend it to their own particular plans. GitLab can have a sustainable business model serving the long tail of companies and developers that want a reasonably affordable hosted service while the top 20% can pursue more specific endeavors using the same technology.

That way we remove the discussion of confidentiality, the way companies treat our data, and we move forward into more productive actions such as building a tool that benefits both the external public and each other’s internal plans.

Walled gardens are here to stay. Releasing integration — while opaque — public APIs is not enough. It’s very exciting to see those new options arising to fill this gap. The OSS environment along with a sustainable business model makes sense. And I hope to continue to see more and more competitors to closed services following the Hybrid model in the near future.

Why go through the trouble to install, maintain and tweak open sourced alternatives to well-established hosted (albeit opaque) services? I believe it’s more than a discussion over convenience, maintenance and cutting costs. We should start small, less risky, but as we grow we should be able to take control back. But we usually can’t get out of walled gardens without a significant - and sometimes impossible - investment in reinventing the wheel. We always had the small open sourced components that those services are built on, but to go from those small components to a full-featured system, it’s an unrealistic path.

With this possible trend of post-SaaS, or Hybrid-SaaS as I call it, we may have just gained the missing link to go from just convenience to full control without the inherent costs and risks chasm.

If you’ve been off traveling around Jupiter for the last few years, you may not have heard about a strand of production that has caused a lot of controversy and already established itself as viable: home 3D printing. 3D printing, or additive manufacturing instead of the subtractive production we use today, has existed for about 30 years. But the market stayed in the hands of giants like 3D Systems and Stratasys, and only for industrial printers that ran past 100 thousand dollars. They were the IBM of the 70s.

And just like the desktop revolution that Microsoft, Compaq, Apple, and Dell helped found, now it’s the turn of startups like MakerBot and Form Labs to step in. Form Labs was one of the most successful projects in its category on Kickstarter. They bring home 3D printers, under 2 thousand dollars, within reach of hobbyists and, down the road, the ordinary consumer.

“But 3D printing is just for nerds.” That’s exactly how “a personal computer in every home” sounded in the late 70s. That’s the premise of the documentary Print the Legend, produced and released by Netflix on September 26, 2014.

I’ll admit it makes me a little nervous that the “3D revolution” already gets a documentary without having reached the level an Apple II had by the mid-80s. Then again, the film doesn’t set out to explain the manufacturing process or the various 3D printing techniques, lasers, and so on. It explores what happens when the garage-startup idealism turns real and hits the level of an early-80s Apple.

The mess is obvious in the interviews with former MakerBot employees, about how they now hate their CEO and co-founder, Bre Pettis. The original garage vision, of a company built around open source and open hardware, suddenly takes on investment and leverage from the usual Brad Felds. And the company starts ballooning from a 12-person garage into a 200-, 600-, and more-person corporation.

Friends turn into enemies (and by the way, friends always turn into the worst enemies), people leave disillusioned, traumatized, disappointed, and cruelty and authoritarianism become routine. At least that’s what the ones who left always say.

How much of this is true? We’ll never know. The truth always has three versions: the one I tell, the one you tell, and the one that actually happened.

People start startups, usually, to be the anti-system, the rebels, the ones who don’t fit in, the ones who want to do things differently and change the game. This level of naive idealism is even “cute,” and it’s necessary: if you don’t believe the thing can happen, it really won’t. Thinking that way is a requirement, up to a point.

I call it “naive” for a reason. In the very rare cases like MakerBot’s (and maybe Form Labs’ down the road, since it’s walking the same path after landing its fat Series A), when you genuinely succeed at changing the system and flipping the game, from that moment on you become the new system, the new game, the new target. Now you’re the one to be fought and torn down.

Apple was the counter-movement to IBM in the 80s. It worked so well that Apple itself became the new system. In the 90s we figured Microsoft would be IBM’s successor, but in the 21st century we found out the one who took on that role was Apple.

It falls to a new garage startup to start another game from scratch. The most interesting part of “Print the Legend” is exactly how it shows that drama, from the beginning until now, of a startup walking that curse. How Bre Pettis went from leader of an open maker community to “traitor” of the movement, and how little MakerBot reached success: it was bought by Stratasys, the second largest in the market, and with that managed to grow the business.

Without giving away too much, someone near the end mentions that founders, entrepreneurs, and successful CEOs always talk about the “sacrifice” it took to get where they are. Most people picture sacrifice as working long hours, staying away from family for long stretches, or grinding through the complications of the process.

But the sacrifice they may be talking about is a different one: breaking your own identity, betraying your own principles, and crossing the lines you swore you’d never cross. And the rare big success stories, the Microsofts, the Apples, the IBMs, and now maybe the Stratasyses, demand that this happen.

The cycle will start over. Soon a new rebel garage startup shows up, plays the same game, and shortly after becomes the new system to be beaten. That’s the cycle of successful startups, and those are exactly the rules of the game of great “success.”

Opening photo: fb.com/printthelegend

Recently a blog article kicked up some buzz. Titled I wanted to work at Apple really bad, and now not so much, it’s a vent from the author himself. The summary is simple: in his opinion, his boss treated him and his team badly, to the point where he dropped everything and walked out.

The more generic discussions ran along the lines of how Apple must really be a horrible place to work, and that the upper ranks still practice the legendary “Steve Jobs style” of kicking people to squeeze out results. Some even went so far as to say “the only way out is to start your own thing, chase your own dreams without a boss like that stepping on you.”

And that’s the point of this article. With the current frenzy around startups, this is an attitude I see more and more often.

Let me put the common complaint about bosses in perspective. No boss you’ve ever run into, even the worst of the worst, comes close to Henry Clay Frick, who ran the operations of legendary Andrew Carnegie’s steel mills. Working conditions in the late 19th century were terrible, and maybe not even the worst conditions today match the worst of that era.

Several of the great industrial magnates of the period, including Carnegie, Rockefeller, Vanderbilt, and many others, exploited workers to the edge of misery. One example of how far it went was the famous Homestead Strike. The result? Seven workers killed and eleven wounded.

Believe me, you would not want someone like Frick as your boss.

That was an extreme example, and it doesn’t mean there aren’t bad bosses today. I’ll keep the context to the world of tech startups we’re part of, and leave out public service, the military, and other areas outside that frame.

The big difference between today and Carnegie’s time is that today there are options. Back then, either you put up with terrible conditions and terrible pay, or you lived in misery with no job at all. Today any good professional has plenty of alternatives. People come and go from companies hundreds of times a day, across hundreds of companies, and you only have to look at the proliferation of job and recruiting sites to see it.

If you’re a student or a fresh graduate, understand a hard truth: you are just like the hundreds of people who graduated alongside you and enter the market every single day. The first job is always hard. And not just the first: the second, the third, and probably the ones after that too. What you “think” you’re capable of is worth nothing until you have a run of concrete results to prove it. That’s what we call a résumé.

I’ve heard plenty of young people say things like “I need to change jobs, I’m not learning anything here anymore.” If you’ve been at the same company for ten years, it’s probably past time. In the first or second year? That’s nonsense. The company has plenty to teach you; the problem is you don’t want to learn.

A programmer, for instance, doesn’t have “programming” as their only job. In August 2013 I wrote that the hardest thing for a Software Engineer is to grasp that software problems often have no solution in software. Learning to negotiate, communicating so others actually understand, managing expectations, working in teams, convincing others that your ideas are good: all of that is part of the job.

In other words, a lot of things that have nothing to do with code. And this holds for any profession, whether you’re a designer, a writer, a cook, or a race car driver. Everyone has to learn and practice these skills to grow.

There are catchphrases repeated thoughtlessly, like “it’s not worth living someone else’s dream as a company employee.” If that’s the thinking, better to resign. A company isn’t a charity: it exists to turn a profit, not to babysit employees. And unless we’re talking about a non-profit, no employee is there as an unpaid volunteer.

Between company and employee there’s a work contract. Both agreed to the terms. If you don’t agree, don’t accept it; if you accepted, honor it; if you think the situation has changed, renegotiate. Renegotiating is an option, and the company can say no. If you’re not satisfied, resign, honor the terms to the end, and leave like any good professional. Walking off the job, like in the blog account, is a unilateral breach of contract. It’s wrong.

You have several reasons to hold a job. There’s the salary to pay the bills at the end of the month, assuming you’ve grown a spine and stopped sponging off mom and dad. And there’s the learning: you’re paid not only to do your job, but to explore everything established companies have already figured out, which you soak up “for free.” Every criticism you get, every complaint, is a chance to sharpen something you didn’t even know you were missing. “No pain, no gain.”

The very fact that you struggle to find another company willing to hire you already shows you’re still short on skills. Another common line is “I’d do more, but nobody gives me a chance (or pays me more).” If you’ve caught yourself saying that, you still have a lot to learn. And even to be an entrepreneur, this is the most basic lesson: nobody gives you anything for free, everything is a voluntary trade for mutual benefit. No benefit, no deal.

And I’ll tell you the worst place in the world to learn all of this: your own first company. The most common outcome when someone completely inexperienced decides to open their own shop? Becoming exactly the kind of boss described by the blog author who worked at Apple.

PS: there were plenty of ways the blog author could have solved his internal problem. A company the size of Apple, with more than 80 thousand employees, surely has several paths to solving a problem like that. Not knowing how to find them only shows there's still a lot to learn. After all, the squeaky wheel gets the grease.

"Everyone's doing B2C startups, almost no one is solving the 'Enterprise' problem, the billion-dollar B2B market."

Let me simplify. I’m treating “B2C,” business to consumer, as all the common sites we see today: social networks, instant messaging, content management, everything whose final target audience is the mass of individual consumers. And “B2B,” business to business, as the software solutions that solve both a corporation’s internal problems and problems that connect multiple corporations, where the target audience is legal entities. It’s a simplification, but it serves my purpose.

I’ll repeat something I tweeted recently: “entrepreneur” is a state, not a career or a profession. Any person, individual or legal entity, by definition, is entrepreneuring. Even a student is, building their own knowledge to solve some problem in the future. A house cleaner is an entrepreneur, and today some even earn at scale by subcontracting others and distributing services.

So it’s not only software developers who, by chance, won this “privilege.” In practice, there’s nothing special about being an entrepreneur, you just have to not be a procrastinator.

Many young people, many not even in college yet, are dazzled by the false “glamour” of this short period, which could well be a bubble, or at least a temporary market anomaly. If you’re in this group, I recommend my previous article, the translation of Alex Payne’s post, Letter to a Young Programmer Considering a Startup.

The exercise here is simple. Why do you think the overwhelming majority of tech startups invest in B2C web applications? And why are almost all of them so similar to each other? Basically, if you swap the logo, the name, and a few colors, many of them come out practically identical. Ecommerces are obvious. Vertical social networks. Mix a network-of-acquaintances component with verticals like jobs, classifieds, transportation (taxi), exercise and health.

The answer is frighteningly simple: it’s all an inexperienced person knows. Their day-to-day boils down to using Facebook, Instagram, Medium, and Tumblr. To taking the bus and the subway. To going to an emergency room or getting some exams. To going to the gym. To attending a university. And all of it from the consumer’s point of view, never the provider’s or producer’s. That’s why everything comes out more or less the same.

The Corporate World Is Very Boring

The most frightening side effect is that “tech startups” and the pseudo-career of “entrepreneur” come as the salvation of whoever read blogs, tweets, and Facebook posts about how ‘horrendous’ the life of a worker at an agency, at a big company, at a corporation is. The bureaucracy, the politicking, the hierarchical structure. All of it painted as if getting a job were walking down death row, and as if being employed at a corporation were being tortured daily in cruel ways.

And worst of all: most young, inexperienced people believe it. TL;DR: it isn’t entirely true.

To make it worse, Black Swans really do show up along the way. Facebook, Tumblr, Snapchat, Twitter. That gives the impression the feat can be repeated. Not understanding the Law of Large Numbers can be fatal.

In short: given millions of people trying to win the lottery, one eventually wins. But there’s no repeatable process that gets you there. It’s a dead-end of randomness. Thank Bernoulli and Poisson for warning us about it, even though few know it today.

I’m not saying there’s no value in B2C. Of course there is, just far more limited than people imagine. The other side of the coin, the corporate world, is almost immune to startups. We know the big names: SAP, Oracle, Microsoft, Totvs. On the purely technical side, this is very low-quality software that, ‘magically’ for some, moves billions of dollars. And it’s a market that has seen little or almost no disruption in decades.

The reason is simple: few have stopped to understand how this world works. Think about someone trying to break into ecommerce. The shallow understanding is that an ecommerce is a “virtual store.” I see it differently: I understand ecommerce as one more distribution channel.

Behind it there are countless consolidated processes that anyone who has worked in retail knows well. Distribution centers, Supply Chain, pricing strategies, taxation and material transport, purchase approval, inventory and warehouse management. Familiar territory for anyone who has worked with Sales and Distribution. Alien to everyone else.

Job sites. Few grasp what that means, but it’s a small leg of the area known as HR, Human Resources. It means understanding labor law, labor liability, benefits, recruitment and selection, training, payroll, timesheets, travel.

Understand this: you have no chance of offering anything to the corporate market without having actively taken part in it. To someone who has never been there, a corporation is a black box. And most immediatist people tend to back away from what they don’t understand instead of exploring it. A “hacker” does the opposite: cracks black boxes open. Whoever just consumes stays a script kid thinking they’re more than they are.

When you pass through several companies, several contexts and situations, different processes at different stages, only then do you start to understand the problem domain. Do this a few times and maybe you start to see how to improve or, better yet, “innovate” in this segment.

Innovation in the corporate market starts with simple process optimization, and can end in the complete replacement of an old process or the creation of an entirely new vertical. But innovation, on its own, is a process, with stages that have to be cleared one by one. And the first stage is invariably acquiring knowledge.

There are no shortcuts. Even in B2C it amazes me to see someone try to build a product for the masses without understanding the persuasion techniques accumulated in the body of knowledge called “marketing.”

The shortcut of building copy-cat tech startups, blindly believing in a nonexistent promise, is raising a generation that’s ignorant of the processes that govern this enormous market. And some things are time-dependent. I took advantage of that by entering the market early, aiming to figure out how its gears work.

When you’re the youngest, mistakes weigh less and wins are overvalued. Once you’re past that age, mistakes become intolerable and wins are treated as the expected norm. And the later it gets, the less chance you have of understanding how the world works, with more and more doors closed.

Obviously I’m generalizing. There are exceptional cases where this works differently, but if you don’t yet know whether you’re an exceptional case, the odds are that you aren’t.

Serendipity only happens when you expose yourself to the widest possible variety of situations, creating opportunities. In a single startup, a single product, a single unproven idea, the chances of serendipity collapse.

The advantage of an SAP, over its several decades of existence, was accumulating the knowledge of every process, of every industry, sweeping through nearly everything important in the corporate world. And even so there are areas it still can’t quite touch, which opens room for a Salesforce.com to stand out. Now picture whoever is artificially limiting themselves.

More than that: with no knowledge at all and the illusion that “mistakes” are good because they turn into learning, we forget the obvious. Erring is human. Erring at what is already notoriously wrong is stupidity, pure and simple.

Charging a client without having signed a contract first, then calling them dishonest for not paying in full, shows the stupidity of whoever didn’t sign the contract. Spending the money up front as “investment” and then running into payroll trouble is amateurism from someone who ignores the basics of cash flow.

Rewriting entire systems without knowing what it will cost in development hours, user training, integrations, data migration, and opportunity cost is another huge blunder from someone who can’t do basic cost analysis. All of this is plain stupidity, you don’t need to err to learn it. Seeing it happen once was enough: every company does this, for decades, long before you were born.

This is inexperience, which is normal, nobody is born knowing everything. That’s why acquiring experience matters so much. Without it you’re always at a disadvantage, because you wanted a shortcut you thought would save 5 years and it’ll leave you 10 years behind.

The B2C world is over-saturated. The B2B world will stay free of important startups for a few more years, since this generation is also wasting time and effort in inefficient ways, throwing away everything we’ve known for years. And remember: avoiding waste, as I recall, is the pillar of “Lean.”

Alex Payne was one of the earliest engineers at Twitter. I know him from Scala and his posts on NoSQL and other tech topics. Beyond being a programmer and a book author, he’s a seasoned startup investor.

His view on startups lines up with mine. I recommend watching the talk, which grew out of a post on his blog with the same title as this article. I decided to translate that post into Portuguese because Brazilians need to hear more of the other side of the coin.

From here on, these are Alex’s words:

I regularly receive emails from young people, usually those interested in programming, who are trying to make decisions about school and/or their professional futures. This post is for those young people.

If you’re in your late teens or early 20s, you’ve grown up in a world that idealized startups, their founders, and the people who work at them.

If you’re in school, you may have felt pressure or have been encouraged to drop out and join or start your own company. If you’re already out in the working world, you may feel that your non-startup job is somehow inadequate, or that you’re missing out on valuable experience and potential wealth.

The generation you grew up in has, in recent years, teetered on the brink of becoming “lost.” Jobs are scarce, entering a university offers no guarantee of employment. Big, old corporations are no longer a guarantee of safe harbor for building a career. Startups appear, through the lens of the media, as the only sign of life in a dying world. I understand the appeal.

Everyone’s background is different, and your choices are only yours. That said, you’re not making decisions in a vacuum. Here are some things to consider that, in my experience, you’re less likely to hear about if you’re working in startups.

A startup is only a means to an end.

I recently interviewed a young guy. I asked him where he saw himself in four years. “Running my own company,” he said without hesitation. I asked him why. “Because entrepreneurship is in my blood,” he answered. There was no mention of what his hypothetical company would do, what problems he would solve for people. His goal was a company just for the sake of having a company. That’s what he studied for, after all.

People are satisfied in their jobs when they operate with a sense of purpose. So much of our understanding of the psychology of success revolves around goals and keeping those goals top of mind: visualizing the moment of achievement, tracking progress, having others hold us accountable to reach them. Goals shape our individual futures.

Maybe a startup is the best way to reach that goal, and maybe it isn’t. If the goal of the young guy described above is to own a business — any business! — then maybe a startup really is the best path. For others, however, I always wonder if they’re adapting their goals into the shape of a startup because that’s the most applauded, admired, and easily understood (if not easily achieved) approach.

Maybe the best way to reach your goal is starting a non-profit organization or going into politics. Maybe it’s doing research at a university or making art. Maybe the answer is joining a large, stable company with formidable resources you can leverage to reach your goal. Maybe your goal is best reached by building a stable, pressure-free career that leaves enough time for your family, friends, community, and self-improvement.

A startup is only a means to an end. Consider the end, and try not to celebrate the means. What do you care about? Who do you want to help? Does a startup make it easier or harder to reach this goal? Where will this leave you when your goal is reached? Where will it leave you if it isn’t?

A job at a startup is the new office job. Startup culture is the new corporate culture.

Startups are portrayed as being exciting, risky, and even subversive alternatives to traditional corporate employment. Startups are seen as freer, more open, more flexible. Some companies really do start like that, but some interviews at late-stage startups make it clear how fast they ossify into structures that resemble the organizations that came before them.

Just as in the first dot-com bubble, there’s currently a proliferation of startups, incubators, accelerators, angel investors, and so on. In order for the “startup community” to replicate itself, nanobot-style, the mechanics of “doing a startup” have been reduced to an easily transportable sequence of actions accompanied by a shared set of values, norms, and language. As a result, the culture of most startups is uniform, even in style of dress and accessories that help “startup guys” — and they’re mostly “the” guys — stay with their herd (so they’re not preyed on by cougars).

Business school graduates like the young guy I interviewed are going straight from university to startups, and that’s if they even finish school. Business graduates, traditionally risk-averse, now say they don’t want to work for big companies. But startups are the new big companies. They are, as I’ll describe below, the office cubicles of a distributed workforce aggregated by venture capitalists and their associated institutions.

There’s nothing inherently wrong with an office job. Just understand what you’re joining. When the company-supplied beer kegs run dry, when the free snacks are making you fat, and playing Xbox in the chill room isn’t as fun anymore, what then? When you find out you now report to a politicking middle manager and not the inspiring CEO who interviewed you, will you still want to be there? Is it an environment with enough toys and novelties to sustain you? When every place you consider working looks more or less the same, how long does the novelty last?

Startups have been systematized, mythologized, culturally and socially de-risked; reduced to formulas and recipes. And yet, there is no sustainable formula for creativity and rebellion. When we try to industrialize innovation, we tear out exactly what we’re trying to cultivate: the creative destruction that keeps the fire of capitalism burning.

Startups are part of the system, not a rebel wrench in the gears.

Funding for startups — that is, the money that pays your salaries — comes from somewhere. Wealthy individuals and institutions invest in startups like any other asset class. The futurist Bruce Sterling recently joked that “startups are full of young people working hard to make other people rich — Baby Boomer financiers, mostly.” Even though this may be a generalist and cynical statement, it’s by no means untrue.

In broad strokes, and excluding areas like biotech, venture-backed startups are a machine where relatively small amounts of capital are inserted on one side and, ideally, a lot comes out the other side. (In reality this machine seems to not work anymore, although the effectiveness of VC as an asset class and the VCs themselves is open to discussion). The salient point: what’s in the middle of the machine is you. You make it run.

The machine isn’t interested in you. In fact, the machine is designed with the understanding that most startups will fail, or at most offer mediocre returns to investors. Most companies in many VC portfolios are recognized as straw men. One or two “10x” companies are what give value to the portfolio. In the best case, founders of failing startups get another push at the slot machine. In the worst case, their failures lead them to despair.

There’s nothing inherently disruptive about a venture-backed startup. The startup system is just another system; an alternative to the corporate ladder but with as many rungs. Some startups may end up dramatically reshaping the market, but so can any other operator already in the game or active regulator.

The now perennial celebration of startup-driven disruption raises the question: if we accept that disruption is indeed happening, are we better off in the resulting disrupted market? Did we solve a problem, or did we just move the problem around? Are we creating value while increasing fairness and equality, creating positive change that will last? Or did we just make one group of rich people a little richer at the cost of another group of rich people? Are we creating a better future, or just reshuffling the present?

Startups have an ongoing interpersonal cost.

No startup story is complete without a scene or two of macho heroes: nights spent programming to exhaustion, weeks away from home trying to raise money in humiliating pitch after humiliating pitch. Startups appeal to a desire for daring that is lost in many forms of modern work, and we hear stories of the rewards waiting when personal lives are sacrificed on the altar of the launch schedule.

Less often we hear about the damage startups do to people’s lives. In extreme cases, like with the entrepreneur whose story of failure and suicide is linked above, the “community” may take a day or two to reflect and lament in blogs and tweets. Then they’re back to “the grind,” and more importantly, making a public show of how hard you’re working to satisfy the investors and the intimidating competition.

I’ve seen firsthand the damage startups can do to relationships. I’ve seen marriages and friendships destroyed, I’ve seen children and partners being sidelined, and I’ve failed those in my life in every way when work came first. I would listen as people who are the picture of a successful startup — visible in the press and social media, leading conferences, eternally founding and exiting — confided their total loneliness despite seemingly being the social center of the entrepreneurial community.

When you’re young, relationships seem like a renewable resource. Friendships exist in abundance and come easily. Just leaving home and into the world, you’re eager to escape the constant presence of family. Being pulled into work makes you feel important, independent, an adult. The work itself provides a new community and new friendships, and the bonds formed during intense collaborative work are strong.

I made many friends through work at early-stage companies. Ironically, these relationships are easier to maintain and enjoy when I’m not working at a startup.

Parting Thoughts.

I’ve worked in startups for half my life now. This work has brought me financial freedom, and I’ve put money I earned at a startup into other startups. If there is, as I’ve argued above, a modern “startup system,” I’m part of it.

I won’t mince words: I am deeply skeptical of this system. I’m skeptical of this system’s slavery, self-congratulatory fetishization of “disruption,” while at the same time becoming so obviously the kind of impassive institution it wants to replace. I’m skeptical of the short-term vision of the startup community. And I’m particularly skeptical of its disregard for the lives of the people who participate in it and the lives of those who live in the world that startups seek to reshape. Let’s not even start to discuss the collusion, cartelization, and other market corruption activities that are so common in the VC world. The point being: it’s a bad game, and a fraudulent game.

And yet. There are startups I wouldn’t want to see disappear. There are people working at and founding these startups who are good, kind, with balanced personal and professional lives, concerned about the impact of their work. Just as we throw slander and accusations of corruption at other systems like politics, mass media, entertainment, and professional sports, we must also admire those who operate ethically and effectively in them. We should celebrate even more those who pioneer new and alternative systems, because they work in the shadows of the community that has a constant hand on the crank of the publicity machine.

Now, you could say that I’m laying too much responsibility at the feet of the startup world. Although this system spreads every day over itself as the savior of everything, from capitalism to culture, of course we can accept that business is business and ideals are better left at home. As a VC from a top firm on Sand Hill Road told me during a pitch many years ago when I was describing a conceptual feature in Simple that would make it easier for users to more easily and regularly donate a portion of their savings to charity, “let’s not waste time on that stuff, we’re here to make money.”

You may be able to stomach that conduct, but I hope your idealism hasn’t been worn down while still so young. I hope you want your work to have meaning, purpose, and value no matter what form that work takes. More than that, I hope you want your life to be defined by more than your work.

Young programmer, I urge you to consider both sides of the startup coin. There are so many ways to leave a mark on the world.

A Note about Feedback

If you are part of the audience of this post, I’d love to hear your feedback and answer your questions, preferably by email. My address is easy to find on this site.

If you are not part of the audience of this post and are really, really angry that a stranger on the Internet has a different opinion from yours, I encourage you to direct your energy into an alternative argument that someone young (or, really, anyone at a career transition point) could benefit from reading. They are your future employees, coworkers, or founders. They are the ones you need to convince, not me.

— May 23, 2013

On different levels, we consider ourselves “rational animals.” We understand our ability to reason as the characteristic that separates us from other animals.

In this world of startups, we spend our days trying to answer a hard question: “How do I demonstrate the value of my product to the world?” Is it the cheapest? The easiest to use? The most elegant?

To get to the root of the problem, it helps to go back a few centuries and then move forward to what we know today about how we reason.

"I think, therefore I am"

Almost everyone has heard and repeated this phrase at some point in life, usually with dubious interpretations, but few stop to reflect on what it means. The phrase is by the French mathematician and philosopher René Descartes, a man who doubted and questioned everything, including his own senses, since they had fooled us countless times. At the very least, the act of questioning, of thinking, defines the only certainty we can have: that we exist.

The name Descartes, in Latin, is “Cartesius,” from which the word “Cartesian” comes. It names Descartes’ philosophy, the “Cartesian Dualism”, the complete separation between mind and body. The idea is that “the intangible soul exists despite the material body.” It’s the entry of mathematical thought into philosophy, with the proposal that all possible knowledge can be deduced purely through reasoning, regardless of the body, the senses, and the emotions.

Let’s jump from the 17th century to the 19th and meet another name: Phineas Gage, a US construction worker still remembered today for an unusual episode. An explosion at a railroad work site launched a huge iron rod, about 3 cm in diameter and weighing 6 kg, which entered under his left jaw, passed behind his left eye, pierced his brain, and exited through the top, leaving a hole in the top of his skull.

The force of the explosion threw the rod 25 meters away. Gage stayed conscious, even while literally leaking brain from the hole.

What makes the case interesting is that Phineas survived and lived another twelve years. According to reports from the time, though, the person before the accident and the one after were completely different.

Those reports aren’t reliable, but they describe the Gage before the accident as “balanced, meticulous, and persistent about his goals, besides being a responsible and skilled professional.” After the accident his personality changed, “he was belligerent, ill-tempered, lazy, and irresponsible. He became impatient and stubborn, using coarse language he had never used before.”

"Phineas was no longer Phineas."

This episode can be considered the catalyst for modern neuroscience and its effects on psychology, sociology, and even behavioral economics. The Portuguese neuroscientist Antonio Damasio uses this case in his famous book “Descartes’ Error”.

"I feel, therefore I am."

Modern neuroscience has already knocked down the old Cartesian notion of separation between mind and body. Today we know that the body influences the mind. For the purpose of this explanation, let’s simplify our emotions as the expression of the physiological state of the body and brain, affected by stimuli. Anyone who has used legal or illegal drugs, Prozac, Zoloft, LSD, or who went through hormonal anomalies, knows the devastating effect this has on their own definition of “SELF.”

Damasio’s hypothesis is that the region damaged in Phineas was the left prefrontal cortex. This is one of the foundations of the “Somatic Markers Hypothesis”, which proposes that our decision-making, far from being purely rational and Cartesian, is guided by emotions.

Research on several people who, like Phineas, suffered damage to the neocortex region shows problems with decision-making and social conduct. They have difficulty planning for the long term, learning from their own mistakes, and behaving in a socially acceptable way.

"A man always has two reasons to do anything: the good reason and the real reason."

This article was born when I stopped to reflect on this phrase, commonly attributed to the finance giant who helped constitute General Electric, John Pierpont Morgan, better known as J.P. Morgan.

Back to the world of biology, nothing and nobody designed us the way we are. We are the most recent result of constant mutations and survival tests in the environment, over millions of years. We are at the top of that chain, Homo Sapiens, and, thanks to thinkers like Descartes, we define ourselves as “rational animals.”

Indeed, one of our most striking characteristics is having a brain equipped with a prefrontal cortex (PFC), which gives us cognitive skills, personality, expression, decision-making capacity, and the moderation of social behavior.

But our brain is the result of evolution, and the PFC is only the most recent episode. According to the triune brain theory, the second oldest part is what we know as the limbic system, the source of emotions, along with the proto-reptilian brain.

A characteristic present in the brains of all animals, from invertebrates up to us, is the instantaneous mechanisms we can summarize as “emotions.” Joy, sadness, fear, courage, pleasure.

It’s a good thing we have emotions like fear, because that’s what makes us react even before reasoning about it. It can be the millisecond of reaction that makes you dodge the car you didn’t consciously notice but that would have killed you. The same millisecond that keeps your hand from getting too close to the mouth of an animal about to bite.

Technically, emotion and instinct aren’t the same thing, but emotion carries that instinctive component.

The important point, and the theory behind Somatic Markers, is that the emotional reaction comes first and only then comes the rational reaction, which takes longer to process. That order matters. Using a metaphor borrowed from Freud’s psychology, we can equate the emotional brain with the “Id” and the rational brain, the PFC, with the “Ego.”

The order we’re used to says that, at least in our slower, more “pondered” decisions, we can “turn off” the Id almost completely and use only the Ego’s cognition. We run cost-benefit analyses, weigh the facts impartially, and make the best possible decision.

What neuroscience shows is that the opposite may be happening. First your Id has already made a decision (“I want that!”) and then your Ego rationalizes and tries to justify that decision (“I want it because of reason X.”). Some people call this “gut feeling,” intuition, a visceral emotional reaction.

That is, the “real reason” comes first, the emotional reaction. Then you justify and express it with the “good reason,” the rational reaction.

This matters because we don’t think in that sequence. Without being rigorous, look at what’s known as the “Marketing Mix,” the 4 Ps of Marketing: Product, Price, Promotion, and Place. In the 1960s, Jerome McCarthy coined the term and Philip Kotler turned it into dogma in his magnum opus, the Principles of Marketing. The 4 Ps already grew into 7 Ps, then 4 Cs, 4 Es, and they keep spawning derivations.

Looking at it from the surface, most of us are conditioned to rationalize: comparing technical specs, trying to justify one product as superior to another because of price, technical features, availability. When we think about how to present our product or service and how to differentiate it, the focus is usually on making it “better,” “faster,” “cheaper,” all rational justifications we imagine will most influence the other “rational animals” like us.

The role of Marketing starts before justifying the product or pleasing the Ego. The primary thing is to provoke the Id first, to create the visceral emotional reaction. The objective features still count, and they become tools for the Ego to use when justifying the choice. But the choice will already be made. That’s the beauty of the theory.

You can justify as much as you want how many goals your team scored this championship, how many titles it won through history, the lineup. All of that is your ventromedial prefrontal cortex trying to rationalize the emotional choice the limbic system has already made. If you love football, no amount of rationalization will make you switch your “heart team.”

You can’t adjust the levels of noradrenaline, serotonin, and dopamine directly, injecting chemicals into people to move their emotions (only psychiatrists can do that legally). But you can affect these reactions through other channels. Human beings have five senses that serve as an interface. Any advertising campaign that features children, for example, wants to manipulate your limbic system into the first emotional reaction before you stop to rationalize.

So far we’ve only scratched the surface of what fields like medicine, neuroscience, psychology, sociology, and economics have been researching in recent decades. The intention is just to show that these fields exist, that knowledge about human behavior evolves all the time, and that you can draw many new perspectives from it to use in the real world.

“We’re putting together an excellent new product that will solve your problems, would you use it?” says the validation survey of a new startup. Result: 70% of the 1,000 people who answered said yes. So the conclusion is clear: the product is valid and can’t go wrong, right?

Wrong. Those numbers don’t mean a thing. On their own they carry no validity, and they support no conclusion at all. “But it’s 1,000 people! How can you say that?” This article isn’t trying to be statistically rigorous, just to point out a few of the many places where almost everyone gets it wrong when running a survey these days.

First of all, what’s your product’s target audience? Let’s say it’s the population of the interior of the southern states, classes C and D, ages 50 to 60, with little experience using computers. Now imagine the question above was part of a survey done online, shared on Facebook and Twitter.

Who answered it? Plenty of people from classes A through C for sure, concentrated in the southeastern capitals, ages 16 to 25, on a computer every day. You asked the wrong question to the wrong people.

For the sake of the example, let’s say you took care to pick the right population and dodged what’s known as “Biased Sampling.” Even so, the question is still bad, and it should be obvious why. It’s what we call a “Loaded Question”: the wording already nudges the respondent toward a certain answer.

For starters, it takes for granted that the product is “excellent” and that it “will solve” the person’s problems. A less loaded form would be: “We’re building a product with features X; assuming you have problems Y, do you think this would help solve them?”

There are plenty of techniques for steering clear of baseless premises, like actually explaining the value proposition instead of stuffing the question with rhetoric. It’s very easy to build a fallacious question that pulls a false answer out of people.

It gets worse when you take those 1,000 people, single out the 700 who said “yes” to the rigged question asked of the wrong audience, and start drawing conclusions from it. “Oh, we noticed that of these 700, half have a credit card, shop online at least once a month, and say they’d share the service with at least 2 friends on social media.” And it gets worse still when the ones who said “no” happen to be the audience closest to the target you want to reach.

I also consulted Ricardo Couto* who explained the issue. “Knowing the respondent, their habits and motivations is much more valuable for the business than having a high percentage of favorable responses about the product or service in an online questionnaire. By the way, calling a form put together in 5 minutes on Google Docs or SurveyMonkey a ‘survey’ is, at best, naïve. What you have there is a questionnaire. Questionnaires can be useful if used well. But in general, they’re just sets of questions around a subject. A survey can even make use of that tool, but not every questionnaire is a survey.”

Statistics, probability, and research in general are the work of professionals who master this field. It’s no trivial matter, and it’s not something to handle carelessly. This example is just one of the many ways to fool yourself with numbers.

Mark Twain popularized the phrase “There are three kinds of lies: lies, damned lies, and statistics.” The subject has been discussed and rebutted so widely that it produced a famous 1954 book by Darrell Huff, “How to Lie with Statistics,” which catalogs the intentional and unintentional errors in interpreting statistics.

Ricardo also added: “If we think of an online questionnaire as a serious research tool, first you have to know how research is done. I really doubt that the people who say they ‘are doing’ a survey and send you a link to a questionnaire took into account the order effect, the observer effect, or avoided leading questions – factors well known to anyone who does real research. And I’m not even getting into statistical validity, which you only reach by applying statistical methods (a simple percentage isn’t descriptive statistics, okay? you have to know what a p-value is), or into selecting respondents suited to the target audience you intend to reach. Or do I only want, as users, customers, or consumers, professionals in my age bracket, my income range, and my online shopping and social media behavior?”

So the next time you go to create an account on one of those dozens of online survey services, check whether the tool meets the bare minimum of a well-done survey and walks you through the right steps: defining the audience, collecting, processing, and analyzing the data rigorously. Without that, all you’re left holding is a pile of opaque numbers, with no meaning and no conclusion to draw from them.

*Ricardo Couto is a Consultant in User Experience and Research, Mentor at Aceleratech, President of the Interaction Designers Association SP (IxDA SP), and Member of the Brazilian Information Design Society. He studied Cognitive Psychology (master's) at the Federal University of Pernambuco and specializations in Information Design, Higher Education Teaching, and Distance Education Technology. He was a university professor at undergraduate and graduate level for 7 years.

In 2007 I wrote two articles on “consultancies,” A Vent and A Vent II. They’re worth a read to understand the context. I like to reread my stuff to see whether I’d still write the same thing today.

The first is a set of tips for a consultant to improve themselves. In the second I describe the greatest sin of a consultancy. I wrote both when I already had a few years at companies in different industries, the most recent being several years at a consultancy.

The first article is very simple, and what I recommended back then is what I still recommend today: continuous improvement. Two points from the original text:

It's no use just changing jobs, for two reasons: a new project won't necessarily bring new things (the chances of being just another form are high). And leaving learning only for when you need it, on a project, is already too late. You'll be overtaken by others who know what you should have already known. It's a good programmer's obligation to know things before, not after: after a terrible quality product goes into production, the result will be bad both for your client and for your reputation.

Stop complaining. Complaining creates a negative mentality, irritates others, and doesn't lead to any solution. Complaining is just another excuse, a noise that hides the real problem. The problem is this: you're complaining about the job, about the project, about the boss, but in reality you'd like to be complaining about yourself. Complaining is just the externalization of the frustration of being what you are and not being able to be more. And that's not a physical inability, it's just mental laziness.

For me, the great advantage of working at a consultancy instead of staying at a single company? It’s being exposed to different projects, different companies, different people, and different ways of dealing with each type of problem.

Look at it this way: a consultant is seen as a “pariah.” A company’s employee, in particular, has no reason to like a consultant, because they’re usually called in to do the employee’s job, in less time, with better quality, and earning more. In practice, if the consultant succeeds it’s because the employees were doing a terrible job before. It’s not always like that, but that’s the general view.

From the consultant’s point of view, it’s a hard job. They have to step into a context the employee had months to absorb, figure out what needs to be done and why it hasn’t been, and produce more than everyone else with less authority and less knowledge. A consultant is effectively a mercenary: we’re called in for the work nobody else could pull off.

Note: anticipating the freedom to interpret the “mercenary” metaphor, let me clarify I mean “the one called in to solve the problem no one else could” and not “being paid to do whatever they’re asked, including the wrong thing.” With the metaphor clarified, let’s move on.

That’s what I like. And that’s why a good consultant can’t regress to the standard-employee mind, the same way a mercenary doesn’t compare themselves to the private soldier: they need to be the soldier and the general. When I was a consultant, what pleased me most was being aware of this position, walking into a client knowing I’d have to solve the problems the insiders wouldn’t.

It was feeling the pressure of walking the front line without knowing what I’d face in the middle of the jungle. Talking to the subordinates, to their bosses, to the surrounding departments, and mapping the terrain to find the best path through the minefields. Being able to disagree directly with the client and, at the same time, offer a better solution than anyone had thought of. That’s the exercise of every good consultant.

The second article, A Vent II, revolves around a problem I’ve seen at many small and medium consultancies: poor financial health. That’s “The” main problem:

When a consultancy shows signs of financial problems, it's the red signal: jump out as soon as possible. Don't think twice. If the client decides to throw a tantrum, if the client decides to delay the invoice, if the client decides to stall, none of that is the consultant's problem: that's exactly what we pay this insurance to the consultancy for — so that it places itself between the client and you. The consultant, correctly working the hours billed, has the right to receive on the dates in the relevant contracts.

A large consultancy, with an established name for at least a decade, can get a sale price (rate to the client) up to 5 times higher than the buy price (rate to the consultant), or more. But that’s the exception in most of the market. In my case, the maximum is 50% of the buy rate, and almost 2/3 of gross revenue goes to payroll.

That’s the reality of the services market. With a tight margin, not counting other operational costs and investments, commercial and financial control is fundamental. You know a consultancy is competent when you’ve never had to complain about your payment, in a sector where clients never pay on time.

There are several advantages to working alongside a good consultancy instead of going it alone as an autonomous worker or freelancer, and being paid on the right dates is one of the main ones. If payment comes in unstable, or turns into a stiff, the freelancer option gets attractive again. So when I decided to start my consultancy, this was one of the main concerns. The rule I never broke: everyone gets what’s in the contract, on the stipulated dates.

And after more than 1 year, one of the most unstable periods in a services company, that rule held intact.

But that’s not all a good consultancy offers. The word “consultancy” is a big umbrella, unfortunately distorted by misuse. I split the definition into at least three big niches.

The first is agencies. Here the consultancy is about marketing and advertising, at most, and the software part usually carries low value in the total budget, sometimes handed over more as a “freebie.” Execution is low quality because the pieces are volatile, temporary, and don’t need to last long. Anything different from that, done in agency style, rarely works well.

The second niche is the infamous software factories. In them, most people work as pure “procedural executors,” never quite reaching “consultant.” The fundamental trait is missing: direct contact between the programmer and the client.

The executor works inside a defined process, receives work orders, clocks a fixed number of hours, and their only task is to deliver what’s already been detailed, in fixed roles. There are variations, but in general the factory is driven by a process that tries to be deterministic. For that, the kinds of software tend to be limited: forms, registration screens, and form flows, especially company intranets or integrations with back office.

The third niche, “consultancies,” where “consultants” actually exist, is rare. The assumption is that the kind of professional at a consultancy is superior to the one at an agency or a factory.

To understand this niche, we need to see what the disadvantage of being a freelancer is. A freelancer has to be a mini-consultancy: doing their own personal marketing, cultivating relationships and a commercial network, negotiating rates, scopes, and deadlines, closing contracts (and most close them verbally, because they don’t know how contracts work), issuing invoices, billing, and so on. Oh, and in the middle of all that they still have to program.

Their compensation tends to be higher because of that overhead, but it’s precisely because of it that productivity and quality tend to drop. And when the overhead isn’t handled well, the result is low-trust deals closed verbally, the way you should never do it.

A good consultancy removes that overhead and does it in a superior way. It’s a service exchange, my ideal concept of “voluntary exchanges for mutual benefit”. The alternative to being at a consultancy is acting as a freelancer or becoming an employee at a company whose core business isn’t necessarily software. And here’s a parenthetical: even at a tech startup, I’d say the core business is rarely the software itself, reflect on that.

“Services” is a free market where exchanges are essentially voluntary: service in exchange for compensation, as it should be.

There are companies where software is the core business, like software consultancies. And others where it isn’t, like commerce companies, where software is accessory and not the main business. Or hosting companies, where the core is unquestionably infrastructure and operations.

That’s another thing I learned as a consultant: identifying the core business. If a company sees software as “profit,” that’s where it invests, that’s where the future is. If it sees software as “cost,” it tends not to invest, and to swap what it has for cheaper alternatives.

And that makes sense. Why invest in software at a company whose core is sales? Or finance? There, software and programmers are cost, like a Word or Excel license. If the goal is to grow professionally, look for companies where software is part of the core. If you want a quiet job, routine and few activities, accepting the invisible risk of going obsolete fast and being subject to cuts, look for companies where the core isn’t software.

Back to the difference with a freelancer: the consultant doesn’t have to worry about marketing, contract negotiation, legal, billing, cash flow, or HR. Like a good mercenary, they get a mission and carry it out. More than that, the consultancy closes projects that would never be within a freelancer’s reach, and offers a delivery security a freelancer can’t provide.

A good consultancy also doesn’t end the contract with the consultant just because a project wrapped up. That’s another freelancer pain point: when a project ends, a sales cycle begins that usually drags. A good project takes 1 to 2 months of negotiation to sign, from first contact through proposals and adjustments. The consultant doesn’t carry that weight; they just deliver with consistency and quality, while the consultancy handles the overhead of keeping a healthy pipeline of projects always coming in.

Now, the key point: “like a mercenary.” That’s the most undefined part of the model. I believe a good consultant doesn’t need an “operational boss,” the traditional “manager” who defines work orders, takes over communication with the client, does the internal politicking, and tells each “employee” what to do.

Of course there are senior and junior consultants, and juniors will always mirror and learn from the seniors, but that doesn’t make them “bosses.” A good consultant is like a good mercenary: either they know what needs to be done, or they figure out how to do it, the right way.

“Autonomy” is the buzzword everyone learned to repeat like a cuckoo clock. Everyone only remembers the “freedom” part of the word, and conveniently forgets the “responsibility” part.

“Oh, but I am responsible.” You aren’t. Your name isn’t on the company’s articles of incorporation. If the company takes a loss, goes under, or has to deal with its own debts, the one who pays is the partner, not the hired consultant. Legally, you’re never responsible, until you become a partner.

And that’s a real possibility for whoever turns into an excellent senior. At my company it exists, and it was recently exercised with the entry of 2 new partners.

If the consultancy gives you the “freedom” to run a project the way a consultant needs to, no intermediaries, no operational boss, no short leash, it’s because there’s far greater trust than any company would give an employee. If something goes wrong, the consultant has zero risk and the consultancy bears 100% of it.

That’s another crucial factor some don’t get. It’s like insurance: would you drive a new car with no coverage? The consultancy is the consultant’s insurance. There they train on projects where they technically have no “responsibility” and still gain knowledge, capability, and networking, at zero risk.

That’s what I liked most about the consultancy where I worked between 2002 and 2007. I came in as a junior programmer, practically a trainee, even though I already had at least 5 years of prior experience. It was all new, so it meant taking a few steps back to learn. And I learned.

On some projects the challenge was purely technical, building software on platforms, tools, or languages I’d never used. On others, it was coordinating the work with colleagues the best way possible. On others, it was talking to the client to find out what they needed and executing the best solution. In the end, the challenge turned into selling, convincing the client of the right solution.

I never held the title of “manager,” “pre-sales,” or “architect,” but I acted in all those roles, switching from project to project as the situation demanded. At one moment I was the only developer assigned for a short stretch to solve a technical problem nobody could. At another, I was the one putting together a team of 10 people, hiring and coordinating. In all of them I did what I knew I should do: solve problems. A consultant who creates problems has stopped being a consultant and become just another employee.

A consultant’s compensation is also different from an employee’s. Part of why employees grow more slowly is that their pay rises almost automatically, nearly independent of individual performance. Even the best-equipped companies, with formal reviews and all, end up at the mercy of the ones who, instead of working toward growth, work to find holes in the system and do less. Public-servant profile, pejoratively: automatic annual raises, scheduled title changes, a bad environment for anyone chasing growth.

A consultant, like a freelancer, grows by performance, by delivery, and by how much the clients themselves recognize they deserve more. I’ve seen plenty of excellent consultants who, on switching consultancies, took the client along because of their exceptional performance. Today I don’t see that anymore, just unethical ex-employees trying to negotiate with the consultancy’s client on the side, in talk only. And what good client falls for the sales pitch of an ex-employee who clearly doesn’t deliver results?

The biggest problem at a consultancy is when the consultant regresses to the standard-employee stage. The symptom: instead of chasing results, they chase excuses not to do the work as they should.

There’s a big difference between the one who complains and keeps delivering good, consistent results, which is what they committed to when signing the service contract with the consultancy, and the one who complains and delivers less and less at lower quality, already going against what was agreed when hired.

The clearest sign shows up in “unionization” movements (in quotes, pejoratively, because unionizing in a market that isn’t monopolized or cartelized makes no sense). It’s the one who produces little and starts stirring up the people around them to create “movements,” smoke screens to pull attention away from themselves. It’s the one who spreads slander, insult, and defamation about colleagues, clients, and the company they work for. Easy to spot: it’s always the one who wants to “speak for you, for the good of the group.”

Exactly because of this, every consultancy has a turnover rate higher than that of ordinary companies, and it needs to. It’s the sector’s dilemma: a services company holds its intellectual capital in its people. Letting go of an employee, much less a consultant, means losing an important asset of work and relationships.

That’s why, when a services company lets someone go, it’s a conscious decision, aimed at admitting a hiring mistake and taking a step back to take two forward right after. Part of this I described in the article Net Negative Producing Programmer.

The only dismissal that actually harms a company is the one that happens after there are already signs of financial weakness, in the form of delayed salaries. Either way, not everyone was born or wants to be a “consultant,” and some only find that out over time. It’s all a matter of perspective.

To start with, a good consultant speaks for themselves and doesn’t need intermediaries. And they should know that everyone knows everyone in this market, where reputation and credibility are currency. Just think: what company would hire someone who defames the places they passed through?

Sometimes the consultant doesn’t understand the advantages of their own position. More autonomy always comes with more responsibility, and more responsibility always brings more challenges and more difficulty. Everyone asks for “more challenges,” but few take them on when they show up and say “let me handle it.”

Many look for a job as an “employee,” in the sense of deferring responsibility (the famous “I just work here”). Technically there’s nothing wrong with that, not everyone was born to be a problem solver. Some prefer to work on one thing, to be around a smaller group of people for longer, to limit their perspectives because they have no interest in cultivating (and maintaining, since knowing isn’t enough) a large network.

What many are after is at least a “false illusion” of “stability,” a word that lost its meaning in the 90s, after outsourcing and globalization. Or a “false illusion” of “purpose,” signing on to an “idea.”

I could write a whole article about “signing on to ideas or ideologies,” but I’ll sum it up like this: mostly it’s a false expectation, from a lack of experience and exposure to a broader perspective. As I said, the advantage of a consultancy is giving you that time of exposure to different ideas, people, situations, and perspectives, which trims the rough edges off illusions and teaches you how the real world works.

The thing is, there’s that annoying factor called “time.” Most young people today have less and less patience and are more impatient for instant results than I was back in my day. The irony is that my seniors used to call me impatient, and even so, for me it was always natural to take one step back to walk two forward.

On those who want to open their own business, many justify themselves with lines like “I’m going to work for me and not for others.” That line is a huge false premise. In practice you “never” worked “for others.” Have you ever worked for free, with no compensation at all?

If you got paid, then by definition you worked for yourself, trading your capacity and the quality of your service for money. When you open your own company, you’ll make the same trade again. The earlier and the more inexperienced you do it, the proportionally greater the chance it won’t work out. It shouldn’t be a hard line to grasp, but understanding comes with knowledge and experience, and not with recklessness.

And what defines a “consultancy” for me? I like to think of it, philosophically, as a “meta-company”: a company that carries parts of other companies and needs to be better than them. It doesn’t mean it always will be better, but that’s the “mission” to be pursued without letup. A career in general is exactly that, and being a consultant is taking on that mission more explicitly.

In the end, my goal was always not to repeat what I described in my 2007 Vent: to build a financially healthy company, win clients and close contracts a freelancer couldn’t, with security for both sides. To create a routine that doesn’t require working more than 8 hours a day and 5 days a week. And to build teams as cells exclusive to each client, without the intermediation of an “operational boss,” giving each consultant the chance to learn all the non-technical functions I mentioned above.

2012 was a year of learning and experimentation, with far more problems than I could foresee. Fortunately, my years in consulting gave me the foundation to close out every one of them. Honestly, I can’t picture myself in 2002, when I started in consulting, handling today’s situations and making the same hard decisions.

If I had opened a consultancy in 2002, or any company, I would have closed it, unless I’d had “luck.” And betting on games of chance isn’t what makes a good consultant.

Happy New Year! See you in 2013.

I wrote about my journey in 6 articles, originally in Brazilian Portuguese. You can use Google Translate that’s available in all of them, I’m curious to see if they work well enough. The journey starts in September 2012 and ends in October 2012. Here are the articles:

• Trip to San Francisco - Part 1 - Techcrunch Disrupt

• Trip to San Francisco - Part 2 - Visiting Companies

• Trip to San Francisco - Part 3 - People and Tourism

• Trip to Europe - Part 1 - Amsterdam, Netherlands

• Trip to Europe - Part 2 - Berlin, Germany

• Trip to Europe - Part 3 - Tel-Aviv, Israel

PS: I was in doubt about considering Israel in “Europe”, seems like it’s actually Asia but it’s considered part of Europe for sports :-)

I also posted hundreds of photos organized in Facebook Albums that you can access here:

• San Francisco
• Amsterdam
• Berlin
• Tel-Aviv/Jerusalem

Finally, I’ve documented my every move while there through Instagram, you can see them all here

Enjoy!

This behavior is what I also call “Amateur Hysteria.” And I say this with total peace of mind because I myself have had moments like this.

First of all, let’s define what is commonly called “Legacy”: basically any code done up until yesterday, especially if you weren’t the one who made it, is considered “Legacy.”

But what most programmers fail to understand is very simple: with that definition, any code will always automatically become “legacy” the minute after it’s finished: including your own.

The neighbor’s garden always looks greener? For programmers, the neighbor’s code always looks worse. How many times have we heard?

• “How could it take all this time to make this code? I would have done it in less time.”
• “How could the code have been made this ugly way? I would have done it better.”

Context, context, context

Why did a certain project take “more time than I think it should have?” Dozens of reasons: departments that didn’t collaborate, poorly defined business rules that changed several times, and so on.

Why isn’t the code as “elegant” or “well done” as “I would do it?” Because it was born with a goal in mind, and that goal changed, the way it always changes. The code piled up, the deadlines tightened, the refactorings didn’t happen the way they should. Technical debt grew despite the warnings of the programmers involved, and the result ended up much worse than it should be.

Another thing that should be obvious: every programmer will always find something they don’t like in any code, whether it’s effectively bad and poorly made or even well-made and well-structured. Software is so complex you can find a flaw in anything. It’s like pointing out flaws in a human being: everyone is imperfect.

The fallacy of whoever is listening is to think that just because someone pointed out a few flaws, everything is automatically bad. It’s the reasoning that goes: “my new programmer said they found this flaw and this other poorly done thing, so the system is bad.” Except you didn’t look at the qualities and didn’t weigh whether the flaws actually outrank them in volume, criticality, and severity.

More than that: any new code will also have points another programmer won’t like. And this cycle is infinite.

Now the truth: the programmer who arrived later and questioned all these points, had they been at the exact same moment when everything started, would have delivered the same bad code as the final result, or even worse.

Fear

Now the challenge is to understand the following. In most cases, “Legacy” code is code that is in production, in use, generating results. Something your “new and elegant” code isn’t doing yet.

Another story that many programmers prefer to ignore is the “Big Rewrite” that was never concluded and never went live. Especially when you were the one who worked on it.

The worst type of software is that which doesn’t generate value. And a factor that differentiates a professional programmer from an amateur is exactly how they deal with others’ code. Talking yesterday with Rodrigo Yoshima, he dropped a pearl: dealing with legacy is for adults, and any child can do “greenfield” (a project done from scratch).

It takes much more technique, much more skill, much more capability, to take a legacy and make it better, even if in some cases it is actually necessary to redo some parts. But that shouldn’t be the first automatic alternative without justification. “Because I would do it better” isn’t justification.

The Environment

The process that facilitates teams generating bad code is always one of the hardest problems to solve. Bosses, clients who ask for “urgent last-minute changes” is one of the symptoms. Departments with different priorities. In environments like this, any new software will quickly be made worse to the point where it seems necessary to redo everything again.

I won’t repeat everything the Agile communities have already said to this day, but the “fix” most of the time doesn’t just involve programming techniques, but rather programmers who can see beyond the code and understand how to help fix the environment around them. This includes the programmer colleagues around, the bosses above, the demands coming from surrounding departments, and all the feedback coming from the market in the outside world.

In the real world, software is made by people for people. Understanding that people aren’t difficult means the challenge for a professional is much greater than following good practices or being specialized in this or that technique. A good professional knows that just complaining and sulking doesn’t lead anywhere.

Opportunities

As always, stories aren’t rules, but they serve to illustrate a point. Years ago I was allocated alone to a big client who had a small “legacy” system that was extremely poorly made. A one-off allocation. And I can say “poorly made” because it effectively didn’t execute as it should. It was technically flawed and returned many errors. It was exactly the kind of code that would make any programmer say, immediately, “throw it out and do it again!”

I had 2 weeks to do something. Others had already tried, the client was already convinced to make a new one. No one would think it bad if I said nothing could be done besides starting from scratch. It was an attempt. And I hate saying I can’t solve a problem.

They needed the current one to work minimally. I was a Java programmer at the time, and the system was one of the worst ASPs I had ever seen up to that point, mixed with several DCOM components that miraculously executed somehow.

In 2 weeks I understood what the business rule they wanted was. I understood why the code didn’t work according to these rules. I understood where the data feeding it came from, how it was handled and stored, what results were expected, and how I should intervene to implement that processing.

I used exactly the same code, did a minimal sanity cleanup, removed what was unnecessary, and inserted the minimum necessary to reach the goals expected by the client within the deadline that was stipulated for me.

During those 2 weeks I understood more about the process, understood more about the client, understood more about the environment. Besides fixing the software I managed to build a good relationship. This good relationship, in a few months, led me to return to work for this client but this time on a 1-and-a-half-year project with a team of almost 10 people, where I could coordinate the development my way.

In this case, it was clear that the code was done in bad faith by a programmer who didn’t really know what they were doing. Many pieces of software are really done by simply bad programmers. But I understand that many cases are also good programmers who had to solve a problem within a series of restrictions they had to deal with.

To me, “legacy” is software done precisely by those who call others’ software “legacy.” For me, “legacy” can often be an opportunity, precisely because few can handle it. And for me, people who judge others’ code without knowledge of the whole context are just another amateur wanting to show off.

Of course, this isn’t a defense that all legacy software is justified and good. Not at all. My point here is focused on the automatic behavior of programmers who use this as an excuse.

This was a long year. Personally, one of the worst and one of the best all at once. One consequence was that I blogged very little, but 2012 should be the year of the comeback.

To close the year on a good note, here’s a thought. It was a year of many new startups and plenty of people willing to chase the opportunities of building a business. Anyone stepping into that world needs to take seriously the philosophy behind any capitalist venture, real capitalism, not this mediocre mixed-capitalism that most people practice. Without a proper philosophy, the business world doesn’t go very far.

I draw my source from articles by Ayn Rand and Leonard Peikoff. As a light entry, let’s look at An Answer for Businessmen, published on May 15, 1962, in an unidentified news magazine found among Ayn Rand’s papers.

To wrap up the introduction, here’s the main read: Why Businessmen Need Philosophy, written by Leonard Peikoff, an excerpt from the book of the same name that anyone interested in entrepreneurship should read.

An Answer for Businessmen

If you want to save capitalism there is only one type of argument that you should adopt, the only one that has ever won in any moral issue: the argument from self-esteem. Check your premises, convince yourself of the rightness of your cause, then fight for capitalism with full, moral certainty.

The world crisis of today is a moral crisis—and nothing less than a moral revolution can resolve it: a moral revolution to sanction and complete the political achievement of the American revolution. We must fight for capitalism, not as a practical issue, not as an economic issue, but, with the most righteous pride, as a moral issue. That is what capitalism deserves, and nothing less will save it.

I should like to suggest that you begin by applying to the realm of ideas the same objective, logical, rational criteria of judgment that you apply to the realm of business. You do not judge business issues by emotional standards—do not do it in regard to ideological issues. You do not build factories by the guidance of your feelings—do not let your feelings guide your political convictions.

Don’t Try to Cheat People in Business

You do not count on men’s stupidity in business, you do not put out an inferior product “because people are too dumb to appreciate the best”—do not do it in political philosophy; do not endorse or propagate ideas which you know to be false, in the hope of appealing to people’s fears, prejudices or ignorance. You do not cheat people in business—do not try to do it in philosophy: the so-called common man is uncommonly perceptive.

You do not doubt your own judgment in business—do not doubt it in the realm of ideology; do not let the unintelligible gibberish of the “liberal” intellectuals intimidate you or discourage you; do not conclude: “It must be deep, because I don’t understand it” or “If this is what intellectual stuff is like, then all ideas are impractical nonsense.” Ideas are the greatest and most crucially practical power on earth.

You do not hire men as heads of your business departments, without firsthand knowledge of the nature of their jobs and of how to judge their performance—do not do it in regard to your public relations department; learn to judge whether the stuff they are selling you is poison or not. You do not hire witch-doctors as mechanics or engineers—do not hire them as P.R.’s.

Know Your Friends and Your Enemies

Know how to tell your friends from your enemies. Know whom to support in philosophical and political issues. If you are unable to speak freely, if you are bound and gagged by the disgraceful injustice of such evils as the antitrust laws—at least, do not praise, spread or support the philosophy of your own destroyers; do not grant them the sanction of the victim. Give some thought to the possibility of establishing a civil liberties union—for businessmen.

And if you still wish to have a “social” mission or purpose—there is no greater service that you can render mankind than by fighting for your own rights and property.

Instead, I’m going back to zero and redefining everything. The first thing I need to correct is the use of the word “Democracy.” It doesn’t describe the kind of organization I have in mind.

My line of thought always started from Darwinian Evolutionism. It’s the only way in which Chaos tends toward a certain Order, through self-organization. This process involves several mechanisms, and it’s about them that I’ve already given talks: Scale-Free Networks, Complex Adaptive Systems, and how all of that leads to processes and methodologies, including the much-discussed Scrum.

This line of study, which you can follow in my blogs and talks of the last three years, brushes past biology, sociology, psychology, philosophy, physics, and mathematics. In the end, the most coherent path seemed to lead to what today is called Democratic Organizations, mainly because of famous cases like Semco, of Ricardo Semler, plus democratic aspects observed in companies like Southwest Airlines, Dreamhost, Groupon, and Zappos.

Except I fell into exactly what I always criticize: a handful of positive evidence about a model, however good it may be, never proves the model. Concluding from positive evidence alone is the same as cargo cult. So I need to correct course: the kind of organization I describe and defend has nothing to do with “Democracy.” In my conception, the term “Democratic Organizations” is a mistake.

Democracy vs Republic

Thank goodness our country is called “Federative Republic of Brazil.” These days I don’t know anyone who speaks ill of the word “democracy.” I won’t give a politics lesson, since I’ve studied almost no political science. Without going into all the ramifications, I want to go down first to the minimal definitions of the term and the original ideas.

I don’t defend the owner of the site I’m about to quote, nor all of his ideas, but the text serves my purpose, so I decided to translate it directly. The original is this article by Gary McLeod. If you search Google for “democracy vs republic,” you’ll find hundreds of articles detailing the differences. For now, this one serves. Here’s the translation:

Rule by Law vs Rule by Majority

Just after the completion and signing of the Constitution, in reply to a woman’s inquiry as to the type of government the Founders had created, Benjamin Franklin said, “A Republic, if you can keep it.”

Not only have we failed to keep it, most don’t even know what it is.

A Republic is representative government ruled by law (the Constitution). A democracy is direct government (Akita’s note: yes, there is representative democracy, but as I said I won’t go down all the ramifications) ruled by the majority (mob rule). A Republic recognizes the inalienable rights of individuals while democracies are only concerned with group wants or needs (the public good). Lawmaking is a slow, deliberate process in our Constitutional Republic requiring approval from the three branches of government, the Supreme Court and individual jurors (jury-nullification). Lawmaking in our unlawful democracy occurs rapidly requiring approval from the whim of the majority as determined by polls and/or voter referendums. A good example of democracy in action is a lynch mob.

Democracies always self-destruct when the non-productive majority realizes that it can vote itself handouts from the productive minority by electing the candidate promising the most benefits from the public treasury. To maintain their power, these candidates must adopt an ever-increasing tax and spend policy to satisfy the ever-increasing desires of the majority. As taxes increase, incentive to produce decreases, causing many of the once productive to drop out and join the non-productive. When there are no longer enough producers to fund the legitimate functions of government and the socialist programs, the democracy will collapse, always to be followed by a Dictatorship.

Even though nearly every politician, teacher, journalist and citizen believes that our Founders created a democracy, it is absolutely not true. The Founders knew full well the differences between a Republic and a Democracy and they repeatedly and emphatically said that they had founded a republic.

Article IV Section 4, of the Constitution “guarantees to every state in this union a Republican form of government”…. Conversely, the word Democracy is not mentioned even once in the Constitution. Madison warned us of the dangers of democracies with these words,

“Democracies have ever been spectacles of turbulence and contention; have ever been found incompatible with personal security or the rights of property; and have in general been as short in their lives as they have been violent in their deaths…”,

“We may define a republic to be … a government which derives all its powers directly or indirectly from the great body of the people, and is administered by persons holding their offices during pleasure for a limited period, or during good behavior. It is essential to such a government that it be derived from the great body of the society, not from an inconsiderable proportion or a favored class of it; otherwise a handful of tyrannical nobles, exercising their oppressions by a delegation of their powers, might aspire to the rank of republicans and claim for their government the honorable title of republic.” James Madison, Federalist No. 10, (1787)

“A wise man will not leave the right to the mercy of chance, nor wish it to prevail through the power of the majority. There is but little virtue in the action of masses of men.” Henry David Thoreau (1817-1862)

Our military training manuals used to contain the correct definitions of Democracy and Republic. The following comes from Training Manual No. 2000-25 published by the War Department, November 30, 1928.

DEMOCRACY:

• A government of the masses.
• Authority derived through mass meeting or any other form of “direct” expression.
• Results in mobocracy.
• Attitude toward property is communistic, negating property rights.
• Attitude toward law is that the will of the majority shall regulate, whether it be based upon deliberation or governed by passion, prejudice, and impulse, without restraint or regard to consequences.
• Results in demagogism, license, agitation, discontent, anarchy.

REPUBLIC:

• Authority is derived through the election by the people of public officials best fitted to represent them.
• Attitude toward law is the administration of justice in accord with fixed principles and established evidence, with a strict regard to consequences.
• A greater number of citizens and extent of territory may be brought within its compass.
• Avoids the dangerous extreme of either tyranny or mobocracy.
• Results in statesmanship, liberty, reason, justice, contentment, and progress.
• Is the “standard form” of government throughout the world.

The manuals containing these definitions were ordered destroyed without explanation about the same time that President Franklin D. Roosevelt (F.D.R.) made private ownership of our lawful money (US Minted Gold Coins) illegal. Shortly after the people turned in their $20 gold coins, the price was increased from $20 per ounce to $35 per ounce. Almost overnight, F.D.R., the most popular president this century (elected 4 times), looted almost half of this nation’s wealth, while convincing the people that it was for their own good. Many of F.D.R.’s policies were suggested by his right hand man, Harry Hopkins, who said,

“Tax and Tax, Spend and Spend, Elect and Elect, because the people are too damn dumb to know the difference.”

It Can’t Be Democracy

If you research at least the Wikipedia articles on Democracy and Republic, you’ll see that the subject is much more complex than this.

What I most admire about the Founding Fathers of the United States is the philosophical and political clarity that led them to describe a Republic based on inalienable Individual Rights. Out of that came solid foundations like the Declaration of Independence and the Bill of Rights. I’m not saying they were perfect people, they all had a thousand flaws, but that’s not where we should get stuck (that would be the Straw Man Fallacy).

The so-called Democratic Organizations fail precisely at this point, in the foundation. It simply doesn’t exist. And worse, from a philosophical point of view, it’s incompatible with the Constitution.

There’s no “official” document with these principles, so I’ll borrow the list from WorldBlu, an organization that tries to spread the idea of Democratic Organizations around the world. For the details, it’s worth reading their page. In summary, the principles are these:

1. Purpose and Vision: normal, any company needs this clear, what its core business is, and what it exists for.
2. Transparency: there are controversies, not everything can be open, including some things that are actually legally protected.
3. Dialogue + Listening: I agree, the concept is valid, and against top-down authoritarianism.
4. Equality + Dignity: hard to define, some differences in treatment are more cultural than mandatory.
5. “Accountability”: hard to translate into Portuguese, but it’s something like “Responsibility”; in this concept it’s more about better defining who is responsible for what.
6. Individual + Collective: this is where the whole concept starts to fall apart.
7. Choice: implementable, but within limits.
8. Integrity: mentions “morally” and “ethically” correct, and that’s a problem.
9. Decentralization: instead of centralization, I’m always in favor of this, but within limits.
10. Reflection + Evaluation: a “touch” of Toyota’s Hansei+Kaizen model.

The last two points, Decentralization and Reflection + Self-Evaluation, are the only ones I would call “principles.” If you read the articles I linked in the introduction, you’ll understand why.

My problem with these principles is that they are rhetorical, and so they don’t work as principles. Saying “it has to be transparent” is pure rhetoric. I can claim I never told a lie, and who’s going to disprove me?

The same goes for “dialogue” and “dignity”: they’re good rhetoric, but they describe conduct. They’re process, not foundation. They don’t answer the real principle: “Why should I act this way?”

And this lands on “Integrity,” on doing what is morally and ethically right. But who defines what is morally and ethically right? Without a philosophy behind it, everything stays undefined and open to interpretation, because there’s no objective, unambiguous principle holding the idea up.

Private Property

I finally get to the points “Individual + Collective” and “Choice,” which deal directly with “Democracy.” These two knock down all the others. Democracy and Collective are bad principles. Any system where the majority can crush the minority, and the minority has no objective, defined property rights, doesn’t work by definition.

And “right,” to be clear, isn’t just anything. Things like “right to bonuses” or “right to more rest hours per week” run into the fundamental problem of any organization: “And who pays the bill?” If, for one individual to have a right, another has to work for free, that’s slavery and, by definition, cannot be a right. Read my article on Rights of Man to understand it better.

More than that: “Transparency” as unrestricted access to all company data, and “Choice” over every relevant matter, should never be automatic rights of employees.

Let’s get to the foundation of all this: every company and organization is private property of its owner, partners, or shareholders. It does not belong to the employees, unless they are also shareholders, for example via a stock options program or something of the sort.

Employees are individuals who trade their capability and their work hours, in a voluntary negotiation with the company, for a value called “salary.” It’s a fair, voluntary exchange, for mutual benefit. The employee’s right ends where what is written in the signed work contract ends. There’s nothing to “claim.”

Want more access to confidential information and more say in strategic decisions? Earn it. Make yourself deserve it, don’t expect it to fall into your lap automatically. Show good work, deliver results, and demonstrate the capacity to decide; only then earn the merit of being part of it. From the employees’ point of view, the principle of an organization should always be meritocracy.

But what about the various companies that are positive “evidence” that “democratic organizations” work?

Here’s the key point of this whole article: these cases of Democratic Organizations are not necessarily Democratic. At least not as people imagine it should be.

Actually, if you read my articles on animal groups, chaos leading to order, and self-organization, it makes full sense within limits. No organization needs to control even the color of the sock each employee wears. That’s not efficient or productive. How the team works, what hours, with what tools and codes of conduct, is something the team itself can decide.

Except that, to get to that point, the company’s goals were already defined: where it wants to go, what it wants to produce, how it wants to enter the market, and all the rest.

What the teams will decide “democratically” is how that strategy gets implemented. The “upper echelon” still exists, the company still belongs to its owners, and the strategy arrives ready. What’s left to the employees is the tactics, the implementation.

This is the natural path of organizations. If you want to call that “democratic,” fine. I prefer to slap no label on it and treat it as one more evolutionary step of “Organizations.”

Leaving “the way of implementing the strategy” in the hands of the people who will implement it is just another way that seems efficient to run a business. It doesn’t necessarily work for all of them. And in the companies treated as cases, I believe we find more evidence that they adopted more efficient means of production. The “democratic” label runs too big for what actually happens in them.

It’s like the Toyota Production System (TPS, or Lean for us) at Toyota. It doesn’t show up on the popular “democratic organization” lists and probably doesn’t even fit WorldBlu’s principles, but it does employ “democratic,” or “participatory,” tactics in the implementation of production. That’s the real goal.

And no, I also don’t think the right term would be “Republican Organizations,” just because I talked so much about Republic at the start. It’s neither one. Organizations are Private Properties that need to be managed. In its foundation, the whole organization cannot be “democratic” to employees who aren’t shareholders, by definition.

A publicly traded company can even have an almost “republican” structure. But the point where we want to study and intervene is the day-to-day management of teams, and there we have a “silo” with “participatory” implementations inside a wrapper that isn’t democratic.

The reason for this long explanation is simple: if you think too long about the colloquial definition of “Democracy,” it quickly slides into “Collectivism,” “Populism,” “Communism,” and “Socialism.” I’m absolutely against this line of evolution, it’s philosophically unhealthy. In the political realm, it should stop at “Republic.” In the business realm, it’s simply “Capitalism,” nothing added or taken away.

And the title “Democratic Organizations Don’t Work” is, of course, a provocation. It points to the fact that I’m correcting myself on the wrong use of “Democracy,” when in reality we’re talking about “Organizations with Efficient Processes,” especially the ones that grow organically with an evolutionary-Darwinist logic, like Lean/TPS.

I love the Wear Sunscreen video. I think I first saw it in 2003 or 2004 and it stuck with me as a warning, something to remember once in a while. The parts I like most:

“Be careful whose advice you buy, but be patient with those who supply it. Advice is a form of nostalgia. Dispensing it is a way of fishing the past from the disposal, wiping it off, painting over the ugly parts and recycling it for more than it’s worth.”

Another important part:

“Maybe you’ll marry, maybe you won’t. Maybe you’ll have children, maybe you won’t. Maybe you’ll divorce at 40, maybe you’ll dance the funky chicken on your 75th wedding anniversary. Whatever you do, don’t congratulate yourself too much, or berate yourself either. Your choices are half chance. So are everybody else’s.”

I was reminded of this while reading What the Dog Saw by Malcolm Gladwell, in the chapter where he talks about another author I like, Nassim Taleb, who wrote the excellent Fooled by Randomness. That chapter happens to be on Gladwell’s blog, and this is the excerpt that caught my eye:

“For Taleb, then, the question of why someone was a success in the financial marketplace was vexing. Taleb could do the math in his head. Suppose that there were ten thousand investment managers out there, which is not an outlandish number, and that every year half of them, entirely by chance, made money and half of them, entirely by chance, lost money. And suppose that every year the losers were tossed out, and the game replayed with those who remained. At the end of five years, there would be 313 people who had made money in every one of those years, and after ten years there would be nine who had made money every single year in a row, all out of pure luck. Niederhoffer, like Buffett and Soros, is a brilliant man. He has a Ph.D. in economics from the University of Chicago. He started the idea that by rigorous mathematical analysis of patterns in the market an investor could identify profitable anomalies. But who’s to say that he isn’t one of those lucky nine? And who’s to say that in the eleventh year Niederhoffer will be one of the unlucky ones, suddenly losing everything, suddenly, as they say on Wall Street, ‘blowing up’?”

The article is long, but I’ll spare you the details: Victor Niederhoffer literally ‘blew up’ in October 1997. He lost practically everything.

The Apple Miracle

In 1997, Apple was considered practically out of the market, with a bad product line, terrible image, declining market share, huge losses. That’s when Steve Jobs returned.

In 2009, Apple is one of the most admired companies in the world, its market value multiplied dozens of times, shows extraordinary profits, the brand is one of the most recognized and adored in the world, its products are envied by the whole industry.

This is the story of the CEO of the Decade, from iMac to iPod, to iPhone and now the iPad. We all love stories like this. Imagining that a great brain, a genius, a hero, alone bends an entire industry to its feet. Twelve uninterrupted years of success.

I love this story too. It has fascinated me since the 90s, and I’ve known the whole arc since it began in 1976: the Macintosh in 1984, Pixar in 1995, the great return in 1997, the iPod in 2001, iTunes in 2003, the iPhone in 2007.

But I can’t help imagining how much of it was pure luck, and when the day of the great turn to bad luck will come. I hope it doesn’t happen, of course, but history shows it’s a question of time.

The Magic Recipes

What I find funniest is how many “analysts,” “specialists,” and “gurus” like to cite stories like Apple’s, Warren Buffett’s, or the Victor Niederhoffers of the world as the keepers of “Magic Recipes for Success.”

There’s an entire industry, from psychologists to MBAs, Ph.D.s, and economists, with tons of books and analyses tracing the routes of great people and successful companies. They all try to find what those people have in common and, from there, sell procedures that “guaranteed” will lead you to success.

It’s an easy task: if it works, the formula works; if it doesn’t, it was you who didn’t follow the formula correctly. You can’t go wrong.

I’m tired of hearing: “we should follow this path, because that’s how company XYZ, of great success, does it.” Little do they know that much of it can be pure luck. And luck is uncontrollable, tied to the situation where it happened, impossible to replicate on demand.

In a chaotic world, it’s impossible to replicate the millions of variables that led to a phenomenon. In retrospect, we can see many of the macro factors, but we will never know all the micro-factors that influenced the result. So it’s futile to try.

You can learn from mistakes. Learning from successes is much harder, especially the kind of success that takes an Apple from near-bankruptcy in 1997 to stardom in 2009. Not even the largest supercomputers in the world would be enough to analyze all the variables involved.

Admiring great successes is a good exercise. It inspires us to try different things and to review our points of view, and there’s something healthy in that.

But it’s stupid to go around quoting or imitating others. “Let’s do X because this successful guru said so.” That’s blindness, it’s acting without knowing what you’re doing, it’s failing to understand that someone’s success may very well have come from pure luck. A lot of people prefer to believe it came from sweat, personal effort, and merit, 100% the fruit of their perspicacity and intelligence.

As Taleb says, this sounds like heresy in a culture that treats “sweat = success,” but that’s how the world works. Species evolve or go extinct based on pure randomness. Homo sapiens is here by pure chance. It’s a fact to accept.

There’s more: retrospective analyses are dangerous. After the fact, it’s simple to trace the path backward and see most of the steps that were taken. The trick is that in the past, at the moment of deciding, nobody knew how the story would unfold.

That’s why you can’t compare today’s critic, who says “oh, but you should have done Y,” with the decision-maker of the time, who didn’t have knowledge of the future. And retrospective analyses don’t work as recipes for the future, because the set of variables and circumstances will be totally different.

Heroic stories are part of our folklore, our culture. I already talked about this in the article The Power of Myth: Redux, and I recommend taking a look if you haven’t read it.

Given that, one of the dangers is good old analysis paralysis, when we want to evaluate every possible variable before making any decision. Accept that you’ll never have them all, and no matter how much you manage to analyze, your chances of getting it right don’t improve that much. As I said before:

“Whatever you do, don’t congratulate yourself too much, or berate yourself either. Your choices are half chance. So are everybody else’s.”

Having said all that, what should a person or company do? “Satisfy customers?” That’s just one aspect of the business. “Deliver good products?” One more aspect. What you do along the way gives you sustainability or not, but don’t confuse things: the purpose is only one.

A marketer will tell you the main thing is “satisfying the customer.” A production engineer, “increasing productivity.” An investor, “increasing the value of the company.” A computer scientist, “creating innovative technologies.” A sociologist, “making a difference in the world.” And so on.

The purpose of a company is “making money sustainably.” Just that.

Note: I’m not saying everything comes only from luck, of course not. I’m saying it’s useless to try to separate one from the other in past cases, and even worse to use it as a model for the future, precisely because of that uncertainty.

In the second, technology is an accessory, treated as “operational cost.” It’s there only to support the business, much of what we call “back office.” For lack of better terms, I’ll call the first kind “tech companies” and the rest “enterpriseys.”

Why am I bringing this up? Because plenty of decisions get made out of context. Choices that fit an enterprisey end up applied at a tech company, and vice versa. That’s the source of a lot of pointless arguments.

I get why a bank would hesitate to swap some of its DB2 for a CouchDB today, say. I also get why a medical company would balk at trading its embedded C programs for something like the .NET micro framework. A few of them do try. It just isn’t the norm, and that doesn’t mean the technologies wouldn’t work.

The same can’t be said of “tech companies.” In that context, reaching for the newest and most advanced should be normal. Better yet: rolling your own technologies should be normal.

The obvious worry is that this turns into something random, disordered, chaotic. It doesn’t. That’s exactly why companies like Google, Microsoft, Novell, RedHat, and plenty of smaller ones keep something resembling a Research & Development department, or at least the habit of researching and experimenting. It’s why they work hard to hire the people on the cutting edge of new tech, which makes no sense at a bank, an insurer, or a trucking company.

This reasoning seems obvious, and it is, yet for some reason I keep seeing people deciding and arguing things out of context. That one really gets on my nerves. Recent open-source technologies make all the sense in the world at a tech company. And having employees contribute to open source projects makes even more sense.

To make the analogy easier, I said that at a “tech company” the core business is technology (duh). Now picture a restaurant, a company whose core business is cooking good food. If I decide out of context, as if the food were just an accessory, I could say: “why don’t we outsource the kitchen and start buying hamburgers from McDonald’s? It cuts operational cost and guarantees delivery in the quantities we need. And the whole market already knows and likes them.”

Worse still would be cooks with this mindset: “nah, I don’t want to test that new ingredient because it’s more work. I’d rather grab the ready-made seasoning at the supermarket.” That’s what I call “scrambled-eggs cooks,” because anybody can make scrambled eggs.

At an enterprisey, most of the work is usually, literally, “form and report development.” That’s what justifies Software Factories and the hiring of “coders,” the classic restaurant worker who just microwaves somebody else’s frozen food.

And I always separate a “coder” from a “developer.” A developer needs the head of a “chef,” a real cook, trying new things, taking chances on new ingredients and new dishes. That’s what sets an award-winning chef apart from the guy who just heats up a grill.

I’m not trying to run down the profession, only to illustrate the concept. The problem shows up when the microwave guy thinks he’s a cook and that what he does is gastronomy. It isn’t.

One caveat: I’m not saying there are no “cooks” in consultancies or enterpriseys. What I’m pointing at is how the company views this kind of service or expense. As an ex-consultant, I know full well there are great minds out there trying to shift the mindset of entire industries. Thoughtworks is a clear example.

So, first of all, look at the context you’re in. If you’re in a real restaurant, you’re expected to be a cook. Don’t assume that acting like a microwave guy is fine, unless you’re trying to run the restaurant into the ground.

Another example from our world: I figure everyone who reads my blog knows Github, one of the most innovative open source repositories around. Wrap your head around this: it was the work of literally 4 programmers, some of them fresh out of college. They could have stuck to whatever the “market” calls “acceptable.”

Read this post of theirs: How we made Github fast. If you aspire to be a “chef,” none of the technologies they list should be new to you: ldirectord, nginx, unicorn, rails, drbd, proxymachine, haproxy, redis, ernie, memcached. Want more? Around the same time, they shipped two technologies of their own, Resque and BERT-RPC. I’ll say it again: not much more than 4 people.

Want more? Remember Phusion Passenger and Ruby Enterprise Edition? Those are two boys who hadn’t even left university. They’re “chefs.”

At a tech company, that’s the goal. At an enterprisey, it isn’t. Where are you?

Like everything Seth Godin writes, this 2005 book is fun to read (in my case, to listen to). Old as it is, it still lays out a marketing model that holds up.

There are two ideas here I like. The first is his beef with marketing built on feature lists and bullet-point checklists. The worst kind is the feature comparison: “lower prices,” “surveys show consumers prefer us,” “we have more performance,” “we have more security,” “we have item X that the other guy doesn’t.” Almost always, that’s bad marketing.

The best marketing tells a story. A consumer doesn’t pick a product because it beats the competition on 8 out of 10 line items; he buys a story. In our IT world, that’s exactly why so many people buy Apple. Apple sells a lifestyle, and price, market share and the rest of the small stuff fall into the background.

The Apple world has its Cult of the Mac. Does that make us fanboys? It does. The same way anyone who buys a Mercedes SUV is a fanboy: if the goal is just getting a passenger from point A to point B, a Mercedes does it no better than a Volkswagen. And if the goal is killing your hunger, a good steak at Rubaiyat feeds you no better than a burger at McDonald’s.

None of this works unless there’s a good product first, sold with a good story instead of a laundry list. Why isn’t that obvious? On a checklist, the iPhone is one of the worst smartphones on the market, more so in its first version. A measly 2MP camera, no Bluetooth audio, no FM radio or digital TV, no dual SIM cards, no swappable battery. On paper, a loser.

Even so, it was already threatening to squeeze the market share of the mighty Nokia. The reason is that the product is excellent at what it sets out to do, and the story behind Apple’s and the Mac’s DNA is very strong.

And let nobody read this as “similar” products at different “prices.” You need, as Seth likes to say, a purple cow: a genuinely great product. That’s the prerequisite. Without it, don’t even bother selling; it’s a waste of time.

The book’s other point should also be obvious: never try to beat the competition by improving on the very same features they have. Losing marketing is the kind that copies what the other guy has and shouts “we’re cheaper,” “we’re faster,” “we have this item they don’t.” Your only shot is being different and telling a story the others don’t have.

That brings me to a small case of my own I don’t think I’ve ever told. When 37signals put Getting Real online for free, I was sure it should be the first piece of material for the Brazilian Rails community. Back then there was barely any Ruby or Rails content in Portuguese, and plenty of people figured I should start with reference guides and tutorials, to help beginners get going.

I bet the other way. I wanted to translate Getting Real precisely because it would set the community’s “tone”: its philosophy, values and pragmatism. That was my wishful thinking. On November 13, 2006 I put out the first call-to-arms and, after great collective work from the community, we wrapped up on March 26, 2007. Not everyone read the book, but I believe it helped inspire Brazil’s first Rails entrepreneurs.

A lot of people don’t get why we choose Ruby on Rails. They waste time building comparison lists to prove anything else is better. And on the facts they’re not wrong: Ruby is one of the slowest interpreters around, Rails is missing plenty of features others have, and it doesn’t support half the commercial databases the competition does.

The thing is, a factual list isn’t enough. Rails is good enough, and excellent where it counts. It sells itself through a story: the community’s luminaries, the startups that grew up on it, the Agile and Pragmatic philosophy that came along for the ride.

Good ideas spread fast. All of us who try to spread ideas are, deep down, marketers. And to close, I’ll hand the floor to Seth Godin on why marketing is too important to be left only to the marketing department:

In classical project management, there’s the image of the Cost × Time × Scope triangle. The objective of traditional mechanisms is to try to keep all of those points fixed and predictable. PMBOK is precisely a body of knowledge with that objective.

Today we know these mechanisms don’t work the way we’d like. The premise of “control” runs into one detail: real control is impossible, and the bigger the project, the more illusory it gets. We are not fortune tellers. We don’t have the power to predict the future. We laugh at the palm reader, yet we’re the first to think we can nail a 2-year project down to the detail.

Fixed-Scope Contracts are a necessity of classical project management, and their objective is a single one: to legally get a neck to squeeze. It’s the naive thinking that if the project goes wrong, it’s the vendor’s fault. Just charge the penalty and you’re done. It looks like a clever win-lose setup.

In practice it’s the definition of lose-lose. By the time you find out the project is heading to disaster, you’ve already lost many months. Even if the vendor pays the penalty, the client is left without the product, has lost time-to-market, and has paid the opportunity cost. Very large companies take longer to feel it, sometimes years, but the bill eventually comes due, and by then it’s too late.

Agility, on the other hand, requires a Negotiable Scope Contract, precisely because collaboration is a two-way street. Some clients think agility means nothing more than extra productivity and a delivery guarantee. Nobody can guarantee delivery, but you can guarantee a drastic reduction in risk.

Instead of waiting a full year to see if the project worked out, the client can tell as early as the first month whether things are heading in the right direction. Reducing risk means, worst case, losing a month or two instead of a year or two. The concept is win-win: I gain and you gain along with me.

That depends on a bond of trust between the parties, a bond that doesn’t need a contract amendment at every move. Plenty of clients balk at this because they have to work too; the process runs on collaboration. That’s the only way to guarantee real success on a project.

When people ask me about this, I strongly recommend reading the article Negotiable Scope Contract, from Improve It, written by none other than Vinícius Teles, one of the greatest XP authorities in Brazil. It’s a long read but worth it, since it breaks down every doubt in detail. He even left a contract template available for anyone who needs something practical.

I’m also asked how to convince a client to follow this model. Unfortunately I don’t think anyone has that answer (otherwise we’d be broadcasting it already). For now the only way is to first master every concept of agility, and I mean all of it, not the surface, because your client will have questions and you’ll need to answer them. It’s genuinely evangelization work.

Another technique is to offer two contracts: one fixed and one negotiable. The difference is that the fixed one will cost at least two or three times more than the negotiable. You need to explain the reasons. Technically, the fixed-scope contract carries much higher risk. In the negotiable one, risks are minimized.

Maybe the client’s biggest worry is the absence of fixed scope: “But if the scope isn’t fixed, doesn’t that mean I get less for what I’m paying?” That takes more argument. The central point is convincing them that neither you nor the client has the faintest idea what is actually needed. By starting the project with whatever is most prioritized, you guarantee that everything that matters gets delivered first.

Many studies show that up to 60% of the software delivered is never used by the client, indicating the monstrous quantity of features that were thought to be important but proved not to be.

The clients most open to this are the ones who value their own money. Smaller companies tend to resist less: they don’t have the time or the money for a six-month circus of “requirements gathering,” they need value delivered fast and at the best cost-benefit. The other day I heard about a consultancy running six-month sprints where the first “sprint” was requirements gathering. Yeah, hilarious :-)

Companies that are too big have too many layers of managers who don’t own the capital and don’t execute, and care only about climbing the ladder. Real results don’t matter to them. For that type, a fixed-scope contract matters even more, because they need someone to pin the failure on.

Explaining agility to that crowd is usually a waste of time. The justification is that traditional managers are needed to keep control. But if the premise is that control is illusory, traditional managers are dispensable. And they really are, but that’s for another article.

Anyway, I know it’s not easy to sell truly agile projects with negotiable scope, but we have to keep trying. I also know it’s hard to keep turning projects down, but if it’s obvious the client just wants a neck to grab, walk away. You don’t need that kind of project, the kind that delivers nothing but headaches.

In terms of customer value, I imagine Apple asked itself something like: “why do people need removable batteries?” The answers vary, but I only see two reasons. Since normal batteries last around 3 or 4 hours, working a full day means carrying at least two of them, especially when you’re traveling or stuck at all-day conferences. The second reason is the poor quality of today’s batteries, which “wear out” fast: one that used to last 3 hours starts lasting 2 or less.

Apple solved this on several fronts. They rethought the whole battery construction process, with engineers and scientists who specialize in battery chemistry and assembly. They started building the batteries in custom shapes that use the available space as efficiently as possible. Making them non-removable cuts the waste of moving parts, bays, and latches, which frees up a much larger area. On top of that, they put intelligence into chips that adjust the charging voltage dynamically, keeping the battery from aging too early.

It’s the kind of detail that can slip by unnoticed, but I wanted to point it out, and I recommend watching their video. What they did was smart. They killed off waste that everyone treated as “normal” in any notebook, and they rethought something already considered mature. They broke a common-sense assumption, the “obviously batteries have to be removable” one, and built a product that delivers more value to the customer. To me, that’s very Lean thinking.

I like to think of Lean as a philosophy and a culture, the same way I think of Agile. It’s not a process, even though plenty of people treat it as one. At the center is continuous flow, and one way to get there is to eliminate waste. Alongside that come other ideas, like Kaizen, continuous improvement. The ultimate goal is to deliver the most value to the customer and strip out whatever doesn’t benefit them directly.

That’s exactly what I see in everything Apple does. Innovation is the result of this kind of thinking, which at the same time produces a leaner, more efficient process. It’s the same logic behind MagSafe, the MultiTouch trackpad, the backlit keyboard, and the tight integration between operating system and hardware, which matters far more than most people realize. It’s what justifies the superior build quality of Apple notebooks. And that’s not fanboyism, it’s fact: the value of an Apple notebook is indisputably higher.

Zara

This kind of thinking reminds me of James Womack’s books. I’m not sure whether it was in Lean Thinking that I read about the Zara case, which I later saw again in a talk on leadership by Mary Poppendieck at Agile 2007. I like this case because, by coincidence, Zara is one of the stores I most enjoy shopping at.

Zara is a big clothing chain. In 2007 at least, they had more than 990 stores worldwide, mostly in Europe, with revenue of 5 billion euros. Nothing small. And they can go from designing a garment to putting it on the shelf in two weeks, that’s their production cycle.

Their premise is simple and obvious: they don’t know what customers want. So they rely on store managers reporting back to headquarters, all the time, what customers are asking for. In one city, say, people might want more red clothing because red is the color of the local team that just won a championship. Zara’s answer is to deliver that as fast as possible, and they pull it off in two days.

To do that, they make small quantities of each item instead of big batches. And, against what most people would call “common sense,” they don’t outsource mass manufacturing to China or other Asian countries: the logistics of fast delivery would cost too much and wouldn’t pay off against the cheaper labor. So they keep factories in Eastern Europe.

As a result, Zara makes around 11,000 items a year, against an industry average of just 3,000. And it sells roughly 85% of everything it produces at full price, while the rest of the industry sells 60%, 70% at most. The difference is the waste of making too much, because forecasts don’t work. You see it all the time in the “clearance sales,” “liquidations,” and “outlets.” Zara ends up with less than 10% of unsold stock, against at least 20% for the competition.

That makes it clear where Zara’s profit comes from: a more efficient production process. Cutting costs on Asian labor doesn’t even enter the equation. It’s one more example of Lean thinking. (Note: unfortunately I don’t think Zara in Brazil follows the same process, lately it’s been all old pieces and little that’s new.)

Zara and Apple didn’t discover any magic formula. Neither did Toyota when it created the famous Toyota Production System, the best example of Lean thinking there is. “Delivering value to the customer” sounds obvious, and most companies think they’re doing it when they really aren’t. Focusing on fast delivery to the customer is genuinely hard, and it’s not something we’re used to.

Standards

That leads me to another thing I picked up from Mary’s talk and Womack’s books. Most companies have “standards,” “policies,” or simply “rules.” The word doesn’t matter; I’ll use “standard” for all of them.

For most managers, “standards” exist so any new employee works just like everyone else. They were made to be followed to the letter.

That’s the dumbest way to think about standards. For Taiichi Ohno, father of the TPS, a standard is something else. A standard is the “as-is”: how things work today. It’s a faithful description of how work is done and produced in the present, whether that’s right or wrong. It captures the current reality without inventing an ideal.

A Lean production leader’s job is to make sure standards don’t stay fixed for too long. If a month later the standard hasn’t changed, your boss could reasonably ask: “did you do any work last month? because the standards haven’t moved.” In a culture of continuous improvement, the team is always hunting for better ways to do the same thing. Standards are “baselines,” starting points meant to be questioned and improved constantly. Today’s standard has to beat yesterday’s.

The moment you think “this is how we’ve always done it, so this is how I’ll do it today and tomorrow,” you’ve created a stagnant, unmotivated, complacent employee, and worse, one who hasn’t grown. An employee who doesn’t grow means a company that doesn’t grow.

Companies that keep departments cranking out standards, processes, and policies to shove down the throats of everyone else are, to me, “dumb,” or at least “blind.” They ignore their own eyes: their employees. At the same time, they signal a culture where nobody should stand out, everyone has to be the same, dress the same, clock in at the same time, eat lunch at the same time, even hit the bathroom on a fixed schedule. It’s a company that prizes the average, my old enemies, the Gaussian curves, the “normal” distributions. A company that prizes, by definition, “mediocrity.”

Lean companies, by contrast, prize meritocracy. They value room for error which, through the proper process of Hansei and Kaizen, leads to continuous improvement. And that, in turn, raises the standards cumulatively.

This post follows the line of the previous one, Advice for Managers. I suggest reading that one too.