GitHub's annual Octoverse report, covering activity through the second half of 2025, found that more than 36 million new developers joined the platform over the year — roughly one new developer joining every second. Total commit activity rose 25.1% year-over-year, crossing close to one billion commits, among the sharpest jumps the report has recorded.
GitHub's own framing attributes much of this growth to AI-assisted coding tools making it easier for newcomers to go from idea to working code without first mastering every syntax detail of a language. That tracks with a broader shift happening across hiring: companies increasingly care less about whether a candidate can write boilerplate from memory and more about whether they can direct an AI tool effectively, debug its output, and understand the underlying system well enough to know when the generated code is wrong.
Despite headlines about other languages rising fast, Python's role in AI and data science work actually strengthened — its contributor base grew 48% year-over-year to 2.6 million, and it remains the default language for machine learning pipelines, data analysis, and research code. For anyone evaluating what to learn next, this is a useful corrective to the idea that newer languages are displacing Python; in the specific domain of AI and data work, Python's ecosystem advantage is still difficult to match.
The report also found that nearly 80% of new repositories created in 2025 used just six languages: Python, JavaScript, TypeScript, Java, C++, and C#. For job seekers outside the AI research niche, that concentration is arguably more useful information than any single trend story — it says that broad employability still runs through a short, well-known list of languages, and depth in one or two of them continues to outweigh breadth across many.
If you're earlier in your career, the takeaway isn't to chase the newest framework every cycle — it's to get genuinely strong in one of those six core languages, learn to work effectively alongside AI coding assistants rather than avoiding them, and build real projects that prove both. Recruiters are increasingly looking for developers who can combine foundational language competence with fluent, judgment-driven use of AI tools, and the gap between candidates who have both and candidates who have neither is becoming one of the clearer signals in technical interviews.