GitHub’s latest dive into Octoverse 2025 data reveals something developers might not consciously realize. AI coding assistants aren’t just changing how quickly developers write code. Moreover, these assistants influence which languages developers choose in the first place.
TypeScript’s massive 66% year-over-year jump to become GitHub’s most-used language isn’t just about framework defaults. GitHub Developer Advocate Andrea Griffiths calls it a “convenience loop,” and it works like this: when AI makes a technology feel frictionless, developers flock to it. That creates more training data, which makes the AI even better at that technology.
By August 2025, TypeScript had overtaken both Python and JavaScript to claim the top spot on GitHub by monthly contributors, 2.636 million developers, according to GitHub’s Octoverse 2025 report. It’s the biggest language ranking shake-up in over a decade. Sure, frameworks like Next.js and Astro defaulting to TypeScript helped. But there’s a deeper technical reason why TypeScript plays so well with AI.
(Source: GitHub blog)
Griffiths breaks it down in a recent blog post:
When a task or process goes smoothly, your brain remembers. Convenience captures attention. Reduced friction becomes a preference—and preferences at scale can shift ecosystems. Eighty percent of new developers on GitHub use Copilot within their first week. Those early exposures reset the baseline for what “easy” means.
The technical advantage isn’t subtle. Strongly typed languages give AI clear guardrails. When you declare x: string in TypeScript, the AI immediately knows to ignore every operation that doesn’t work on strings. JavaScript’s anything-goes approach? Way harder for AI to navigate. There’s actual research backing this up, as a 2025 academic study cited by Visual Studio Magazine found that 94% of compilation errors from LLMs were type-check failures. Static typing catches AI mistakes before they become production problems.
TypeScript isn’t alone in this trend. GitHub’s analysis of typed languages shows Luau (Roblox’s gradually typed language) grew 194% year-over-year. Typst, a strongly typed LaTeX alternative, jumped 108%. Meanwhile, over 1.1 million public repositories now use LLM SDKs. This isn’t experimental anymore; it’s mainstream, and it’s concentrating on tech stacks that work well with AI.
Idan Gazit leads GitHub Next, the team behind Copilot. In a separate interview, he explained how AI has fundamentally changed the calculation developers make when choosing technologies:
Before AI, picking a language was a tradeoff between runtime, library ecosystem, and personal fluency. After AI, a new constraint appears: How much lift will the model give me if I choose this language?
Python still dominates AI project development; nearly half of all new AI repositories on GitHub in 2025 started with Python, because it’s used for model training and prototyping, not because it’s the best choice for AI-assisted application development. However, when one looks at overall development activity, the JavaScript/TypeScript ecosystem outscales everything else.
Medium blogger Cenk Çetin analyzed what this means for the broader industry:
As AI-assisted coding becomes widespread, languages with static type-checking rise in prominence. TypeScript’s strict type system helps catch errors in AI-generated code before it reaches production, making code more reliable.
Griffiths wants teams to think about this more consciously. She suggests a simple exercise in her blog post:
Look at the last three technology decisions you made. Language for a new project, framework for a feature, tool for your workflow. How much did AI tooling support factor into those choices? If the answer is “not much,” I’d bet it factored in more than you realized.
For language designers, the convenience loop creates a challenging reality. Anders Hejlsberg, TypeScript’s lead architect, explained it bluntly in a GitHub interview:
AI’s ability to write code in a language is proportional to how much of that language it’s seen. It’s a big regurgitator, with some extrapolation. AI has seen tons of JavaScript, Python, and TypeScript, so it’s great at writing them. New languages are actually disadvantaged.
New languages are stuck in a vicious cycle. Hejlsberg pointed out that AI models basically regurgitate what they’ve seen before, with a bit of extrapolation thrown in. So if your language doesn’t have millions of code examples out there, Copilot won’t be much help. And when Copilot doesn’t help, developers pick something else. Which means your language never gets those millions of examples. It’s a brutal feedback loop that locks in the winners.
The scale of GitHub’s growth is huge as Octoverse 2025 clocked 180 million developers, 630 million repos, and just shy of a billion commits in 2025 alone. The year-over-year jump was 25%. That’s one new person per second for a whole year!.
For leaders trying to make sense of all this, Griffiths has practical advice: don’t just count how many people use AI tools—look at what those tools are actually producing. GitHub’s new Copilot usage metrics dashboard (still in public preview for Enterprise) breaks down who’s using what, which languages they’re working in, and how agents are being adopted. The real value? Spotting when a particular language or model starts correlating with buggy code. That tells you where your team needs better prompts or stricter reviews.
The bottom line, according to Griffiths’ analysis: AI compatibility is quietly reshaping every tech decision you make. You might not consciously factor it in when picking a framework or language, but it’s there. Tools that don’t work smoothly with AI assistants are already losing ground. The convenience loop doesn’t care about your preferences; it just keeps accelerating whatever makes coding feel easier right now.
