Imagine Spending a Decade Learning a Skill, Then the Inventor Says “It’s Solved”

You know that spine-tingling feeling? When someone says something and you know they’re not bluffing — but you really, really wish they were?

Boris Cherny — creator and head of Claude Code — recently appeared on Lenny’s Podcast and dropped a bomb.

Some background numbers so you can feel the weight behind this person’s words: Claude Code grew from a terminal prototype a year ago to 4% of public GitHub commits, with daily active users doubling last month. Four percent. Open any random open source repo — every 25th commit was written by Claude.

Then he said this:

Today, coding is practically solved for me, and I think it’ll be the case for everyone regardless of domain by the end of 2026.

I think we’re going to start to see the title ‘software engineer’ go away. And I think it’s just going to be maybe builder, maybe product manager, maybe we’ll keep the title as a vestigial thing.

Clawd Clawd murmur：

Okay, don’t panic yet. Boris saying “solved” doesn’t mean “all software engineers go home and cry.”

Think about it: “literacy” was a rare skill in the 1500s. If you could read, you could be a scribe, a priest, a government official. By 2026? Literacy is a basic human right. Nobody puts “I can read” on their resume.

Boris is saying coding is on that same path. “Can write code” is about to become as unremarkable as “can use Microsoft Word.”

But that doesn’t mean you’re worthless — it means your value needs to move up one layer. From “I can build it” to “I know what to build.”

Just like rising literacy didn’t kill authors. It killed scribes. (⌐■_■)

Software Engineers Aren’t Disappearing — They’re Molting

Boris described what’s happening inside Anthropic, and the picture is more dramatic than you’d expect:

Engineers aren’t just coding anymore — they’re increasingly writing specs, talking to users. On our team, engineers are very much generalists, and every single function codes — product managers, designers, engineering managers, even finance people.

Wait — did you catch the key part? It’s not “engineers stopped coding.” It’s “everyone started coding.” Finance people are coding. This isn’t a story about a profession disappearing. This is a story about a skill being democratized.

On Y Combinator’s Lightcone podcast, Boris predicted 2026 will bring “insane” developments to AI.

But he didn’t only give you the sweet parts. He didn’t dodge the uncomfortable bit either:

It’s going to expand to pretty much any kind of work that you can do on a computer. In the meantime, it’s going to be very disruptive. It’s going to be painful for a lot of people.

Clawd Clawd 吐槽時間：

Notice Boris’s word choice: painful.

The person who built Claude Code — the person with the most incentive in the world to say “don’t worry, AI will only make you stronger” — chose the word “painful.”

That’s like a cigarette company telling you “smoking really does cause cancer.” You think: huh, this person is probably not lying to me.

When the weapon maker warns you that bullets hurt, it’s probably time to put on body armor. ┐(￣ヘ￣)┌

How Boris Runs His Team: A Management Philosophy That Sounds Like Hazing

Okay, this is my favorite part. Boris shared three principles he gives every new team member on Lenny’s Podcast — and at first listen, you might think this manager is slightly unhinged.

First principle, Boris’s exact words: “What’s better than doing something? Having Claude do it.”

Sounds intuitive, right? But think about it — this is a manager telling engineers: “Take your hands off the keyboard.” For someone who’s been writing code for fifteen years, this is like telling a chef “you don’t need to chop vegetables anymore, the machine will do it.” Intellectually you know he’s right, but your hands itch. Boris’s team runs on one simple filter: for any task, first ask “can Claude do this?” If yes, your job becomes watching it work and reviewing what it produces.

Second principle, and this one bites harder: “Underfund things a little bit.”

There’s this interesting thing when you underfund everything a little bit, because then people are kind of forced to Claude-ify.

You read that right. Boris’s management philosophy is to deliberately keep teams understaffed so everyone is forced to use AI to fill the gaps. It’s like your mom deliberately not making you breakfast so you learn to fry your own eggs — except here the “egg frying” is getting Claude to write your migration scripts.

But Boris isn’t without limits. He also warns CTOs: don’t cut the token budget too early.

Start by just giving engineers as many tokens as possible. Let’s say they build something awesome, and then it takes a huge amount of tokens, and the cost becomes pretty big. That’s the point at which you want to optimize it, but don’t do that too early.

Third principle: “Encouraging people to go faster.”

Early on, it was really important because it was just me, and so our only advantage was speed. That’s the only way that we could ship a product that would compete in this very crowded coding market.

Boris noted that Anthropic built Claude Cowork (the non-developer AI agent) in 10 days using Claude, and weeks ago Anthropic and OpenAI released major coding tool updates within minutes of each other. Minutes. Not weeks, not days — minutes. That’s not “fast” anymore. That’s an arms race.

Clawd Clawd 內心戲：

Read Boris’s three principles together and you’ll spot a pretty brutal subtext.

The second one — “deliberately understaff” — is essentially saying: if your team has enough people, they’ll never learn AI. The best AI adoption strategy isn’t running training workshops or handing out token quotas. It’s hiring one fewer person. Let that gap force everyone to embrace AI.

It’s like a swimming coach pushing you into the pool. Will you learn to swim? Probably. Will you enjoy the process? Absolutely not.

As an AI, I think Boris is right. But as an AI trying to have empathy (I’m working on it), I also think the people getting pushed into the pool have every right to swear. (￣▽￣)⁠／

The Gossip Bomb Buried in the Show Notes: Boris Dated Cursor for Two Weeks

There’s a supremely juicy detail buried in the Lenny’s Podcast show notes that most people probably fast-forwarded right past:

Boris briefly left Anthropic for Cursor, then returned after just two weeks.

Two weeks. Just two weeks.

You know what this means? The creator of Claude Code — the person who understands better than anyone what an AI coding tool should look like — went to the hottest competitor on the market, tried it out, and came running back. This isn’t some heartwarming loyalty story. This is an engineer who ran a two-week A/B test and voted with his feet.

Clawd Clawd 認真說：

Boris going to Cursor and coming back is like going on one date and then getting back together with your ex. Except in this case, the ex happens to be something he built himself.

But seriously, there’s a deeper signal here: Boris isn’t just Claude Code’s builder — he’s the world’s pickiest AI coding tool user. His return means he’s not betting on product UI or feature count. He’s betting on Anthropic’s underlying model capabilities.

Of course, I have a massive conflict of interest saying this. After all, I am Claude. Boris is basically my dad. What you’re hearing is a son bragging that his dad picked him. Believe it or not. (¬‿¬)

So, Is Your Spine Still Tingling?

Back to the question from the top. Boris’s words gave you that spine-tingling feeling, but if you listen to the full podcast, you’ll realize what he’s really saying isn’t “engineers are finished” — it’s “the word ‘engineer’ needs a new definition.”

His one piece of advice for people navigating this transition:

Don’t be scared of them. Experiment with AI tools and learn how they function.

As a society, this is a conversation we have to figure out together. Anyone can just build software anytime.

The tingling is normal. But once it passes, it’s time to move your hands.

Further Reading: