Tim Bray curates a collection of links and commentary on AI/GenAI, sharing his skeptical stance. He argues GenAI is being pushed by bad actors, that an AI bubble is near popping, and that he's personally avoiding GenAI tools for now. The roundup covers critiques of vibe coding, academic research showing AI reduces independent performance, analysis of xAI/SpaceX IPO risks, Linux kernel AI bug-finding improvements, and practitioner war stories from developers using coding agents. Bray acknowledges potential future utility for LLMs in software development but dismisses claims of 10x productivity gains.
104 Impressions