The integration of AI models into developer workflows has led to a bias towards specific technologies, stifling the adoption of new tools due to cutoffs in training data. Popular models like ChatGPT and Claude show a strong preference for React and Tailwind, influencing tech choices. Testing shows that these models consistently generate projects using React and Tailwind, while other models like DeepSeek offer more varied options. The bias of AI models towards certain technologies has a significant impact on developer decision-making, potentially shaping the future of software development. More transparent documentation of biases in AI models could help mitigate this issue.
https://vale.rocks/posts/ai-is-stifling-tech-adoption