A quarter of startups in Y Combinator's latest batch, W25, have 95% of their codebases generated by artificial intelligence (AI), according to Jared Friedman, managing partner at YC. This revelation has sparked a heated discussion on the role of AI in coding, with many experts weighing in on the implications of this trend.

Friedman emphasized that the 95% figure only accounts for code written by humans versus AI, excluding code written to import libraries. He stressed that the founders of these startups are highly technical and capable of building their products from scratch, but have instead opted to leverage AI-generated code. This shift in approach has significant implications for the future of coding and the role of developers.

The concept of "vibe coding" was introduced by Andrej Karpathy, former head of AI at Tesla and ex-researcher at OpenAI, to describe the process of coding using large language models (LLMs) without focusing on code itself. This approach has gained traction, with many developers and VCs excited about the potential of AI-powered coding. Startups such as Bolt.new, Codeium, Cursor, Lovable, and Magic have raised hundreds of millions of dollars in funding in the last 12 months, further validating the trend.

However, experts also caution that AI-generated code is far from perfect. Studies have shown that AI-generated code can insert security flaws, cause outages, or make mistakes, requiring developers to debug heavily. YC's Diana Hu emphasized the importance of developers having the skills to read and judge the quality of AI-generated code, stating that "you have to have the taste and enough training to know that an LLM is spitting bad stuff or good stuff."

Garry Tan, CEO of YC, also highlighted the need for founders to have classical coding training to sustain products in the long run. He warned that relying solely on AI-generated code may not be sufficient to ensure the longevity of a product, citing the limitations of current reasoning models in debugging.

Despite these concerns, Tan remains optimistic about the future of AI-powered coding, stating that "this isn't a fad. This isn't going away. This is the dominant way to code. And if you are not doing it, you might just be left behind." As the industry continues to evolve, it will be interesting to see how developers and startups adapt to the changing landscape of coding.

The rise of AI-generated code raises important questions about the role of developers in the future of coding. Will AI-powered coding tools augment human capabilities or replace them? How will the industry address the potential pitfalls of relying on AI-generated code? As the trend continues to gain momentum, these questions will need to be addressed to ensure the long-term sustainability of the tech industry.