• How Google was able to nurture Gmail within the company in 2004, and why it would be impossible today

• The PR crisis that Gmail caused

• How Paul Buchheit knew Bill Gates was wrong about the internet — as an intern in 1995

• Why it’s culturally impossible for Boeing to compete with SpaceX

• How Gmail succeeded despite not being the first mover in browser-based email

• Why Paul is optimistic about San Francisco’s future

• Paul’s view on the future of AI

---

When Paul Buchheit launched Gmail in 2004, it didn’t just change Google, or the technology industry — though it certainly changed both. It changed the world.

Zoomers will find this impossible to believe, but in the early 2000s people were still deleting email. The habit was partly a holdover from our analog culture, but mostly due to space. When Yahoo launched Mail in 1997, one of the most popular email services in the country, users were limited to 4MB of free storage. That’s about the size of an iPhone photo.

Today, the concept of digital storage for human memory is kind of just the oxygen we breathe. Where did I say we were going last week? “Danny, pull it up.” But twenty years ago, Gmail launched a revolution with a status-shattering proposition: “unlimited email.” In practice, Google was only offering 1GB of free storage. But this was something like 250 to 500x what competitors were offering, and it’s hard to overstate how crazy it seemed at the time — to the extent that many people thought the product was an April Fools’ Day prank. Gone were the days of sorting email into folders, where they were quickly lost to memory, or trashing them completely. Gmail was designed for searching all your old correspondences, documents, notes, which unlocked a new way of working. Now, it was possible to query your old thoughts.

How could something so transformative be built inside a company so sclerotic as Google? It must have been a very different place in the early 2000s. But if the company really was so different at the time, and that different culture enabled the success of products like Gmail, why would the company allow innovation-crippling sclerosis to set in? I had a million questions, and no obvious answers. So I reached out to the architect of Gmail, the legendary hacker Paul Buchheit. [Editor’s Note: check out the full transcript, partially edited for clarity, on Pirate Wires]

“Incentives,” he explained.

Gatekeepers, which comprise the overwhelming majority of employees (and probably people), are incentivized to stasis, risk-averse to the point of paralysis, and terrified — sometimes for good reason — of change. Insurgents are basically insane by society’s standards, anathema to the status quo, and often destructive. But they’re the source of everything new in the world. From the moment a company is founded, always by insurgents, incentives to short-term survival attract gatekeepers, who naturally fear and try to isolate new radicals, then tirelessly work to grow the team of more gatekeepers. On a long enough time horizon, the gatekeepers always take the company. Then, it dies. Gmail, a firebomb happily tossed at the status quo, was fundamentally an insurgent product. It could not have been built at Google even ten years later.

The question for insurgents is how to beat the gatekeepers back long enough to build something new, and with it a lifeboat to keep us all alive a little longer.

Paul has worked in the technology industry since he was a teenager in the 1990s, tinkering with prototypical web-based email. He was an engineer at Microsoft, Intel, Google (employee 23, credited along with Amit Patel — who coincidentally shared the name of Paul’s childhood hero — as the men responsible for the company’s famous former motto, “Don’t be Evil”), and Facebook. Today, he’s an investor, both as an angel and in his capacity as a Group Partner at Y Combinator. We talked about the secret to Gmail’s success, the early days of Microsoft and Google, Paul’s insight into the primal war of bureaucratic gatekeepers against society’s insurgents that has decayed our country to its core, and a little hope for builders moving forward.

***

For years people wondered what would happen if Google — with 190,000 employees at its peak — fired half its team. As in, would literally anything change? To be fair, the question could have been asked of any large tech company at the height of the last bull run, and anyway it was just a funny thought experiment. Or, it was just a funny thought experiment until Elon Musk bought Twitter, cut the team by 80%, and catalyzed a wave of product launches from a company formerly infamous for product paralysis. Suddenly, in April 2022, over a decade after the company was founded, and with a fraction of its former team, Twitter was innovating.

How was this possible?

Paul laughed, as if remembering an old friend — the story of Twitter was insane, a total historical aberration that should not have been possible. But once the company was taken private, nothing that followed was especially shocking.

In any mature company, most employees aren’t building. Most employees are managing builders. They’re keeping track, or count, or carrying messages between the few people working. They’re talking to people, okay? They’re setting guardrails and safeguards and drafting reports.

“They’re gatekeeping,” Paul explained, whereas any real builder is inherently an insurgent. In the context of Twitter, the ratio of gatekeepers to insurgents had so tilted in favor of the gatekeepers that most (now former) employees no longer believed change was possible. Even while it was happening.

“People said ‘Elon should have been more careful’ instead of firing 80 percent of the team,” Paul said. “And everyone thought the service would collapse. They were giving their goodbyes. ‘Oh, farewell! Elon's such a moron. He doesn't know how software works!’ Twitter had a few minor outages, but who cares? It turns out most of this stuff wasn't that important.

“Gatekeepers are one hundred percent anchored to stopping bad things from happening, and they have no concept that when you stop bad things from happening, you are inherently stopping good things from happening as well. You can't ever deliver something that's 100 percent good. If you deliver 80 percent good, that's pretty good. But if you try for 100% — if you try to be perfect — you get nothing. Innovation is inherently not clean.”

I recalled the conversation between Elon and Parag Agrawal, the (very briefly) former CEO of Twitter, right after Elon became a board member, and right before he took the company private. In a lengthy text, with not only his company but the entire industry rapidly shifting beneath his feet, Parag asked Elon to stop tweeting critically about the product. It was bad for morale, he explained. Elon simply asked Parag what he got done that week.

“That’s a great example of the gatekeeper versus the insurgent,” Paul said. “Parag had kind of grown up in that organization. His job was just to keep the wheels on. He’s just trying to keep things from blowing up. Then along comes Elon, who's throwing bombs. Right? He's one hundred percent insurgent. Like, he just shows up and sets things on fire. The first thing he did when he took over was set everything on fire. Then he says, ‘all right, guys, we better move fast or the company's fucked.’”

Community notes, paid verification, subscriptions — it’s worth noting how many of Elon’s earliest new products were actually developed under Jack. But they were gatekept from the public for years. Safety, caution, patience… stagnation. This is the fate of every “mature” technology company. And that word “mature,” Paul said, is really just “a flag for inability to change.”

It’s not as if your average gatekeeper is an idiot, or wants his company to die. Some amount of caution is important, and especially at a company like Google, which is “probably one of the greatest businesses that has ever existed, a goldmine,” Paul said. Naturally, then, the incentives “are basically ‘don’t blow up the goldmine.’” But after decades building caution into the corporate culture, it can’t move. In the words of one senior engineer at the company, “it’s impossible to ship good products at Google.”

“In a gatekeeper organization,” Paul said, “if they have insurgents, they put them in a separate building far away, and they call it like an innovation center or something. Containment. At one point, Google had this thing called Area 120. Every big company that sucks creates an ‘innovation center,’ where they isolate the insurgents in a corner somewhere they won't cause harm.”

Unfortunately for Google, while failure to ship new product in a space like social media never really mattered, failure to ship new product in AI, with competitors in the space threatening the company’s core search product, is existentially dangerous. Five years ago paralysis was fine, maybe even smart (if not exactly wise). This is no longer the case. Given Sundar’s regular lip service to the words “innovation” and “artificial intelligence,” leadership seems to understand the stakes. But it’s too late, the gatekeepers already won — years ago.

Paul reflected back on his experience at the company, right around the time he launched Gmail.

“You had to run through this checklist before every product launch,” he said. “Naturally, the checklist only grew, because it was an opportunity for some ambitious person, within the bureaucracy, to insert themselves into the launch process. Now you need to get a sign off from marketing, or comms, or some other group. Do you remember widgets?”

He kept laughing, through every story, as if he couldn’t believe how crazy it was, the concept of a technology company that kept its engineers from building new technology.

Famously, Google’s widgets team forced the company to use their product by capturing the launch checklist. Once they’d planted their flag, you had to create a widget for your product before you could release it to the public. It’s a perfect portrait of bureaucracy, with all sides fighting not to build, but to control.

On the other side of all of this is the insurgent.

These guys, Paul explained, “are essentially people who want to disrupt the status quo. They want to create something new. Steve Jobs is a great example of the insurgent. He hated IBM, right? They were like Goliath. Think back to his commercial, 1984, when they launched the Mac. Like that's one hundred percent insurgency. Literally, an insurgent runs up and smashes the screen. Steve always had that in him.”

Mark, Elon, Jeff — these are today’s insurgent kings. But every hacker tinkering, and building, or just doing something truly new is inherently insurgent. Their work is always met with skepticism first, and then resistance. Once they ship, however, the world changes. Not because it wants to change, but because it has to change.

“The Boeing Starliner finally launched yesterday,” Paul said. “And of course, after so many failures, it barely made it to the space station. But if it wasn’t for SpaceX? The story would just be, well, space is very hard, experts agree…

“You know, experts studied rocket reusability. It wasn't like no one ever thought to reuse a rocket. Experts studied it, and they concluded it didn't make sense. Like ULA (United Launch Alliance) built this whole model that said it would never pay for itself. If it weren't for SpaceX, they would still just be making up all these excuses, and experts would say, ‘space is just fundamentally hard… It's not that Boeing sucks, it's just that space is really hard.’”

But SpaceX landed rockets, and now landing rockets is table stakes. As the world naturally bends toward gatekeeping, this is the challenge before all insurgents: remove the roadblocks where you can, work with an insurgent, or build your own insurgent company. Then innovate as quickly as possible. Your success alters the world, which is the only way to survive. Then again —

“To me,” Paul said, “part of the joy of insurgency is just blowing shit up.”

Which brings us back to Gmail.

***

“I was really into the internet,” Paul laughed.

It seems obvious now, but in 1995 the notion that the internet would alter our entire reality was not considered common sense. Then, Paul was not a common kind of guy. As a teenager, he didn’t have many friends. He was a Science Olympiad, and didn’t hate spending time with the other kids at Math League, but the thing he most enjoyed was building stuff. He taught himself to code, and rebuilt the controls on an industrial pick-and-place machine, replacing the original control hardware with a 386 PC, running software he wrote.

He listened to the Cure, Depeche Mode, U2, Pink Floyd, and whatever CDs he could grab at BMG/Columbia house for a penny. He liked pizza and coke.

He watched a lot of Pinky and the Brain.

When I asked him about his heroes, Paul said he read biographies of Edison and Tesla growing up, and he wanted to be an inventor. He liked Amit Patel, who wrote a BBS door game called Solar Realms Elite. And then, of course, as was the case with every nerd in 1995 who craved a model for success in the world of computers beyond the alpha business Wall Street archetype, he thought Bill Gates was pretty cool. So when he landed an internship at Microsoft at age 18? That was a big deal.

“At the end of the summer, you got to go to a barbecue at Bill Gates' house, and that was the first time I met him.”

The interns swarmed the man, and conversation quickly turned to the internet, as this — like crypto a few years ago, or AI today — was all anyone wanted to talk about. As Paul recalls, Bill said “it’s good for nerds, or technical people like us. We love it. But normal people want a more curated experience.”

Microsoft was actually building an internet competitor at the time, which would function more like a newsstand where people could safely find and read TIME magazine or whatever. “It was going to be this much more contained, sanitized experience,” Paul explained. Bill didn’t think people wanted any of “these weird URLs, HTTP colon, slash, slash, slash, www dot — you know? It’s like, ‘what is this gobbledygook? A normal person's never going to get any of that stuff.’

“But to me,” Paul said, “what was really cool was it’s Bill Gates saying this, and I'm thinking shit, this guy's the wealthiest man in the world, but I know more about the future than he does. Because he's obviously wrong. The internet's going to fucking blow this stuff away.

Long story short, Paul was not invited back to Microsoft after telling their Internet Explorer lead his ideas were bad, and their product had a glaring, almost offensively obvious security hole. But this was for the best. Really, what Paul wanted was to build a web interface for email, which would free it from your desktop. That’s when he started learning more about the problem, and hacking around, and a few years later he found a team as weird as him.

The Google guys were building on Linux, which is partly why Paul took the job. It was open source, totally decentralized, and deeply insurgent. Google’s founders were builders before they were businessmen, and Larry Page was fully committed to beating back the gatekeepers. “Processes are things you need kill,” he used to say — a joke about unix processes, but also an insurgent’s battlecry at what increasingly felt a lethargic bureaucracy — and in 2001 he reorganized the company.

Google was too slow, so Larry fired all the project managers, and transitioned to “this kind of free range engineer model,” which paired engineers with a VP, and gave them each a product. By then, webmail already existed. Yahoo and Hotmail were the big competitors. But they totally sucked. Paul set out to build “a first class experience inside the browser.”

He’d been thinking about it for years, so version one only took about a day to hack together. He asked for some feedback, hacked a little more, and carried on like this through launch. Later, his critics considered the job messy. They were right.

Gmail was basically a Frankenstein, or at least this is how one “expert” reviewer described Paul’s work. “Because it’s a webpage,” Paul explained, “but it behaves like an application. I also threw in some Unix. So the keyboard shortcuts are from VI, which is a sort of obscure Unix text editor. Why is J K up and down?”

He laughed. It didn’t matter that it was messy, it worked, and his critics never shipped a product as important. In fact, most never shipped a product at all. Paul’s philosophy of work was pretty standard hacker stuff, which is the only way things get built — you simply build. No process, no guardrails. Just do it. Perfection was never Paul’s intention. His intention was to ship, and there was Gmail in the wild. Next question.

Paul continued to improve his work, and Gmail gradually grew in popularity, both within the company and beyond. Finally, much as rocketry was forced to evolve after SpaceX, the industry was forced to respond to Google. At least for a moment, the gatekeepers were defeated, and the company progressed.

Partly, Larry is the lesson here.

A sufficiently empowered founder seems to be the only way to neutralize gatekeepers, and protect insurgents long enough to pull off miracles. Still, even with a deeply committed Larry on your side the odds are never in your favor. Gatekeeping is the natural order. The human impulse to safety is powerful, and probably rooted somewhere in our lizard brain — that instinct to fight or flight that kept our ancestors alive. Progress sounds fantastic, but it isn’t natural. What’s natural is to freeze to death up north in winter, not to build shelter, and furnaces, and power plants, all of which present any number of potential risks. Yet, we build. Or, insurgents resist the mob long enough to build, and humanity progresses.

But the gatekeeper’s pull is powerful. Impossible to resist forever, and not just at a company.

In conversation with Liv Boeree, Samo Burja recently described human civilization as a kind of dynamic, multi-dimensional pyramid scheme, in which elite human beings must regularly contribute new knowledge and energy to the system or all of this collapses. Peter Thiel has spoken of the existential danger in stagnation for decades, not because stasis is dangerous but because it is impossible. Stagnation is an illusion. We grow or we die, and our planet is a museum of civilizations gatekept to their extinction.

It’s a grim question, but worth asking: are we improving our world today, or are we living in the ruins of a world that died with our grandparents?

The last moon landing took place on December 14th, 1972. Today, there are ninety-three nuclear reactors in the United States. We’ve built 3 since the year 2000. Following a century of complications, and a full decade building, New York City recently opened three new Subway stations along Second Avenue. There is presently no city in America capable of building a complete underground rail system. Fortunately, that knowledge hasn’t yet been lost. We’ve just been gatekept to paralysis. But a century from now, with no course correction, who will be left to teach our grandkids what to build, or how?

Technology is, fortunately, naturally corrosive of the status quo, something I wrote about at length in The Fifth Estate. This makes the technology industry an interesting aberration in the world, where gatekept companies tend to quickly die, and startups tend to rise — only to themselves be gatekept to their death. But the cycle continues. In this narrow sense, are tech business a place where insurgents win? I mean, look at Twitter.

“The advertisers nuked as much revenue as they could,” I said to Paul. “But Elon took the company private. He still has to figure out a new model, but provided it doesn’t go out of business the only way to stop him now is some kind of Congressional action, which would be extremely hard to do under the First Amendment.”

Paul was not convinced.

“If Biden wins,” he said, “I don't know. What they're doing with Trump is super scary. Now anyone with high surface area, anyone who gets a lot of attention, can be guilty of an accounting error or whatever, right? If that's the standard, they could put Elon in prison.”

If this sounds crazy, you haven’t been paying attention.

The danger for any insurgent founder is not only, or not even especially, the gatekeepers within his company. Society is full of gatekeepers, and in America they have the power of our courts. From 2021 alone, Elon has faced: a SolarCity acquisition lawsuit, a series of bogus racial discrimination lawsuits, a Dogecoin racketeering lawsuit, a FOIA lawsuit, the Twitter acquisition lawsuit, and one incredibly wild judicial decision out of Delaware to strip him of his previously agreed upon compensation package at Tesla for, I don’t know, vibes?

Altogether, how many lawsuits would you guess Tesla has faced in its short lifetime? By my count: 7,460, almost all of them frivolous. Our courts incentivize this behavior.

In terms of potential future lawfare, the unfortunate truth is anything is possible. Expect a shit ton of Elon indictments, and a judge who simply hates the guy — from that deep, instinctual place of the gatekeeper. Because the difficulty with our world beyond the walls of tech is institutions like our government can be gatekept for entire generations before they finally collapse. The American West is gone, and with it the concept of a political startup. We are all at the mercy of encroaching sclerosis, now, and literally tens of millions of government gatekeepers determined to grind our world to a safe and antiseptic halt. From here, where can we even go?

“This is the constant libertarian question,” Paul said. “It’s why we ended up with concepts like seasteading. You’re talking about the loss of the frontier. I get sort of depressed about this question… ”

He trailed off. But then, very quickly, he seemed to have a change of heart.

“I am actually feeling sort of optimistic about the San Francisco stuff,” he said.

The danger of San Francisco is not only that it’s crazy, but that “crazy” merely starts in San Francisco. Then, it spreads. However, if the city is our path to change, maybe fixing it is also contagious. People pattern match. What happens if Garry Tan and crew really turn the ship around? Could Americans pattern match along a new way forward?

“Part of the thing that's exciting about an insurgent product is people can no longer debate with SpaceX landing rockets,” Paul continued. “People can no longer say it's not possible. You just say fucker, it's right there. I've got a video of it happening. Or Gmail. For years, all those email administrators said ‘you don't understand. It's very hard to offer 30 megabytes of storage.’ Okay, now Google is doing it for free with a thousand.”

What if the same could be done with cities?

Unfortunately, there is no crypto libertarian trick, here. There is no island utopia where you can hide. The rest of the world is as gatekept as America or worse. So our choices are to stay and fight or vanish, along with everyone we love, and everything we care about. But if we win? If we win a single city, even? If we really fixed one place, it might just force a broader change — and it really would be poetic if that happened in the world’s new capital of artificial intelligence.

Paul and I touched on AI a few times, circling the industry’s latest, open-ended vision for the future like Bill and Paul did all those years ago with the internet. AI constitutes the automation of the very work of technology: thinking, building, self-correcting. Now, before the project’s even fully shipped, the gatekeepers have gone to war, as if they understand as well as Paul that this is it, this is the last real battle.

“AI is the most disruptive thing we've ever done,” Paul said, “basically since the invention of fire. It’s like fire, and then it’s AI.”

Google was born in a culture of decentralization, only to gradually centralize and stagnate. But the potential of AI, while an echo of the industry’s past, is significantly different. If we are looking at, potentially, the last human invention, we are also looking at the end of the industry’s natural defense against the gatekeeper.

“AI either centralizes all the power or it decentralizes power. That's the future. We either centralize and we're fucked, or we decentralize and actually it's really great.”

This is not the first time in tech history that gatekeepers have colluded with the government to control a new technology. It might however be the last. Fortunately, for now, the future still belongs to insurgents building. But the battle for our future will be ferocious, and decisive. May the best man win — and may he do it fast.

—SOLANA

--

EDITOR’S NOTE: Check out the full transcript of Solana’s conversation with Paul, partially edited for clarity, on Pirate Wires.