AI Drives TypeScript Dominance, Global Growth, and Secure Development
TL;DR
- TypeScript's rise over Python signifies a shift towards strongly typed languages for building scalable, long-lived applications, driven by AI's influence and the need for compile-time error detection.
- Generative AI and agentic workflows are becoming standard engineering practices, enabling developers to integrate AI capabilities with single lines of code and automate complex tasks.
- The increasing adoption of AI tools like GitHub Copilot empowers new developers, accelerating learning and prototyping, and democratizing software creation for a broader audience.
- Older languages like COBOL and Fortran are seeing renewed relevance through AI's ability to understand and generate code, extending the lifespan of critical legacy systems.
- Global software development is expanding rapidly, with countries like India, Brazil, and Indonesia showing significant growth, fueled by economic factors and government upskilling initiatives.
- Open-source contributions are surging, particularly in AI infrastructure and tooling, indicating a growing reliance on community-driven development for cutting-edge technology.
- Security is shifting towards a "secure by default" paradigm, with tools like Dependabot and CodeQL automating vulnerability detection and resolution to maintain development velocity.
Deep Dive
The 2025 Octoverse report reveals a significant shift in software development, driven by the increasing integration of AI and the maturation of strongly typed languages like TypeScript. This evolution is not only reshaping how software is built but also expanding access to development, leading to explosive growth in the developer community and open-source contributions. The implications suggest a future where AI is a baseline engineering skill, open-source ecosystems become more robust, and global participation in software development continues to rise.
The most striking trend is TypeScript's ascent, overtaking Python in popularity on GitHub. This shift is attributed to TypeScript's strong typing, which prevents runtime errors by catching issues during compilation--a critical advantage for long-lived, large-scale projects with multiple contributors. JavaScript, previously dominant for smaller projects, lacked this robustness. The rise of AI also favors strongly typed languages, as AI models and their associated ecosystems are increasingly built with them. Frameworks now scaffold TypeScript by default, and large language models (LLMs) themselves often default to generating TypeScript code. This creates a powerful flywheel effect: more developers use TypeScript, leading to more resources and tools, which in turn attracts more developers.
Python, while no longer the most popular language overall, maintains a strong foothold, particularly in AI projects. The report indicates that Python dominates AI-tagged repositories and shows significant growth in notebook usage. This highlights a bifurcation: TypeScript is becoming the go-to for full-stack development and large-scale applications, while Python remains crucial for AI research and development. Intriguingly, older languages like COBOL are experiencing a resurgence, not through new development, but through AI's ability to understand and maintain legacy code. LLMs are proving adept at deciphering and updating these foundational systems, extending their lifespan and utility.
The widespread adoption of AI, particularly "vibe coding" and agentic workflows, is fundamentally changing the developer experience and expanding participation. Developers are increasingly using AI tools like Copilot not just for code generation but as collaborative partners, enabling them to prototype and build applications more rapidly. This has led to an exponential increase in new developers joining GitHub--over 180 million tracked, with more than one joining every second. This accessibility means that individuals who might not have considered themselves programmers can now bring ideas to fruition. The implication is that AI is democratizing development, lowering the barrier to entry and fostering innovation across a broader spectrum of users.
This influx of new developers and the increasing complexity of software, now often built on extensive open-source dependencies, underscore the critical importance of security and community practices. While AI streamlines development, it also introduces new considerations. The report notes that security is shifting from a "shift-left" approach to "secure by default," with tools like GitHub's Dependabot and CodeQL playing a vital role in automatically detecting and resolving vulnerabilities. Organizations need to embrace these tools and understand their application to maintain development velocity while mitigating risks. Furthermore, the growth of open-source projects highlights the need for community management, including codes of conduct, to resolve conflicts and ensure sustainable development.
The Octoverse report also points to a significant geographical diversification of software development. Countries like India, Brazil, and Indonesia are showing remarkable growth, fueled by economic expansion, government initiatives in AI upskilling, and the increasing adoption of technology hubs. This global participation suggests that access to the internet and development tools has created a worldwide talent pool, challenging traditional centers of software development and offering new opportunities for remote work and economic growth. The report's projections indicate continued leadership from these regions, particularly India, in shaping the future of software development.
Action Items
- Audit AI integration: For 5-10 non-AI repositories, assess the ease of embedding AI SDKs and identify common patterns for adding AI agents or models.
- Implement agentic workflow testing: For 3 core projects, define and automate agentic workflows that spin up VMs, build tests, and self-correct code based on results.
- Create secure-by-default repository templates: Draft a template for new projects that includes pre-configured CodeQL, Dependabot, and secret scanning, ensuring security is enabled from inception.
- Measure TypeScript adoption impact: For 3-5 existing JavaScript projects, analyze the correlation between TypeScript adoption and a reduction in runtime errors or compile-time issue detection.
- Develop AI-assisted Cobol understanding: For 2 legacy Cobol systems, experiment with LLMs to generate code, identify vulnerabilities, and add new functionality, documenting the process.
Key Quotes
"You know, we see these trends over time, you know, different languages, you know, years and years ago it was C, then C++, then Java, then then Python, etcetera, and now we're in one of those new sea changes where the types of software we're writing is changing, the applications that we're building, they need different things, and so languages appear and disappear based on the functionality they provide."
Jeff Luszcz explains that programming language popularity shifts are driven by evolving software needs. This highlights how the demands of new application types, such as those related to AI and SaaS, necessitate different language features, leading to the rise and fall of languages over time.
"And so what we see right now with the movement to, basically, the internet, software as a service, artificial intelligence, and other things like that, we see TypeScript having its moment. And there's things, why TypeScript? I think it is really having its moment is, um, it's in its name, it's a strongly typed language, so things that might cause problems at runtime, you might have a crash, you might have, uh, basically a hang if you're not doing the right things. TypeScript has operaties in its language that prevents that."
Jeff Luszcz points out that TypeScript's current prominence is due to its nature as a strongly typed language. This feature allows it to catch potential runtime errors during the development phase, preventing crashes and improving software stability, which is crucial for modern applications.
"And when you ask the large language model, please generate me some code, it's probably going to defer to TypeScript. It's the modern language, it, it solves a bunch of the problems that you don't want to have to struggle with as a developer, so things get better with time, and the more people write TypeScript, the more people write TypeScript. It's kind of one of those, uh, those flywheel effects."
Jeff Luszcz suggests that Large Language Models (LLMs) often default to generating TypeScript code. This indicates that TypeScript is perceived as a modern, efficient language that simplifies development by addressing common pain points, contributing to its increasing adoption through a self-reinforcing cycle.
"I think what we're seeing is people voting with their feet, they're, they're looking to upskill around the AI functionality, code generation, agentic flows, especially this upcoming year. And I think we will see people go through this flow: they make an account, they learn, they pick a language, they pick a technology, they, they turn the crank. I'm going to do this, I'm going to build this, I'm going to learn this. So learning how to use these tools, I think also learning how to work with the community."
Jeff Luszcz advises developers to focus on upskilling in AI functionality, code generation, and agentic workflows. He emphasizes that practical application, learning new technologies, and engaging with the open-source community are key strategies for career growth in the current landscape.
"The report mentioned that generative AI and AI workflows are now considered as ordinary engineering. This is really amazing. A year or two ago, when we saw people working on AI, it was really AI experts building new models. They were, 'I'm going to be the next, I'm going to be the next model builder,' 'I'm going to be the next, uh, company with their new model.' And what we've seen this year is it's not just the, the, the, the kind of rarefied atmosphere of like super AI experts, it's people who are saying, 'I'm going to integrate AI into my existing product.'"
Jeff Luszcz highlights a significant shift where generative AI and AI workflows are now viewed as standard engineering practices. He contrasts this with the past, when AI development was confined to specialized experts, noting that integration into existing products is now more accessible.
"For me, I think it's just the number of new developers on the platform every year. I just think that is an incredible number. Seeing tens of millions of new software developers basically being created every year is just kind of mind-blowing and unbelievable to me. I love to see it."
Jeff Luszcz expresses his astonishment at the sheer volume of new developers joining the platform annually. He views this continuous influx of millions of new software developers each year as a remarkable and inspiring trend for the industry.
Resources
External Resources
Books
- "The GitHub Podcast" - Mentioned as the platform for the discussion.
Articles & Papers
- "TypeScript’s Takeover, AI’s Lift-Off: Inside the 2025 Octoverse Report" (The GitHub Podcast) - The episode title, indicating the primary subject of discussion.
People
- Andrea Griffith - Co-host of The GitHub Podcast.
- Kedasha Kerr - Co-host of The GitHub Podcast.
- Jeff Luszcz - Guest, described as a "data whisperer" and one of the creators of the GitHub Octoverse report.
Organizations & Institutions
- GitHub - The platform hosting the podcast and producing the Octoverse report.
- Columbia - Mentioned in relation to Andrea Griffith's role.
Other Resources
- Octoverse report - GitHub's annual report on software development trends.
- Vibe coding - A concept discussed in relation to AI's influence on software development.
- Agentic workflows - A method of building products using AI, where AI agents perform tasks autonomously.
- Code QL - A security offering from GitHub for repository security.
- Dependabot - A GitHub tool that discovers and manages open-source vulnerabilities.
- Secret scanning - A GitHub security offering.