AI Code Generation Favors Mature Languages, Says Arvid Kahl

Arvid Kahl shares a key insight about AI code generation: it exhibits a strong Lindy effect, meaning that older, more mature languages are far easier for AI to work with, while newer inventions struggle to be adopted. He also notes with relief that he no longer has to keep up with new JavaScript frameworks every week, a sign that the era of rapid framework churn may be fading. This perspective matters for developers and tool builders because it suggests investing in stable, well-established languages could yield better results with AI assistance, while chasing novelty might be counterproductive. It also hints that AI coding assistants are likely to reinforce the dominance of proven technologies, making the Lindy effect even stronger over time.