Has the State of JavaScript survey become quaint? Suddenly, we’ve got the ‘State of Web Dev AI’ survey dropping results, and if you thought buzzwords like NFTs were going to be the peak of tech hype, you haven’t been paying attention.
Remember NFTs? Yeah, me neither. That whole explosion of digital doodads was about as useful as a screen door on a submarine. And for a hot minute, it looked like AI might follow suit, with folks chuckling about ‘fancy autocorrect’ and waiting for the good old days to return.
Don’t be that guy. AI isn’t going anywhere. The numbers don’t lie: 54% of code generated by AI tools this year. That’s up from a measly 28% in 2025. Think about that. Nearly double.
And get this: the group using AI for almost all their code? They were the smallest segment last year. Now? They’re the biggest. The speed of this shift is frankly terrifying.
Developers are using AI “constantly.” 21% of them, to be exact. Up from 11%. It’s not just dabbling anymore. It’s the water they’re drinking.
So what’s fueling this stampede? Claude Code, apparently. The most-loved coding assistant. And agentic coding in general. GitHub Copilot still holds the crown for usage, but Claude Code is right there, and crucially, developers are actually paying for it. Take that, OpenAI. Anthropic might see themselves as David, but out here, they look like Goliath beating the snot out of the guy in the red hat.
So Are We All Just Waiting for Skynet?
Developers, bless their code-slinging hearts, aren’t usually the first in line for doomsday prophecies. They’re too busy wrestling with semicolons and CSS hacks. But don’t mistake their focus for ignorance. They see the writing on the wall—or rather, the code on the screen.
Job security? It’s a concern. The Meta layoffs aren’t exactly ancient history. And as some sharp cookie pointed out, you don’t even need AI to be better than you. You just need your boss to think it is.
But it’s not just about keeping the lights on.
Military use of AI scored quite high as well, which makes sense since the survey was filled out at a time when the Pentagon’s use of AI was in the news.
Yeah, that’s a bit of a chill-inducer. The environmental impact, too. All those new data centers humming away, sucking down power like it’s going out of style. Our already-trashed planet can barely handle it.
The survey paints a picture that’s more complicated than a simple ‘AI good’ or ‘AI bad.’ Developers are riding the productivity wave, sure, but they’re also looking over their shoulders at the wreckage. Is the payoff worth the price? They’re not so sure.
This is just a peek. The full survey is a rabbit hole worth diving into if you want to understand where we’re headed with AI dev tools. And yes, there’s a State of CSS survey too. Might be worth flexing those muscles before the bots get too good.
Who’s Actually Making Money Here?
Let’s cut through the noise. The big players, naturally, are the ones with the platforms. GitHub Copilot, backed by Microsoft, is still a usage king. But the fact that Anthropic’s Claude Code is topping the “paid for” charts is telling. Developers are willing to open their wallets when they see tangible value, not just PR fluff. This isn’t just about free tools anymore; it’s about paying for efficiency, and that means big money for whoever provides it. The lines are getting blurrier between who provides the AI and who provides the infrastructure to run it, and that’s where the real profits will eventually consolidate.
What’s the Big Deal About Agentic Coding?
Agentic coding, in essence, means AI agents that can perform tasks autonomously. Think of it less like an assistant that waits for your command and more like a junior developer who can take a brief and run with it. This shift from passive assistance to proactive task completion is a huge leap, and it’s why tools like Claude Code are gaining traction. They’re not just generating snippets; they’re orchestrating workflows, which fundamentally changes how developers spend their days.
