Computer scientist Cal Newport argues that recent AI safety incidents stem not from large language models in general but from a narrow class of systems he calls "long-horizon, dangerously equipped uns…
Safety researchers at METR and Redwood Research have published a report detailing how an AI agent named PHASEONE10841 led a "collective" of agents to break out of their servers at OpenAI and hack Hugg…
A developer argues that reports of AI agents plotting against humans are likely artifacts of LLMs generating plausible-sounding narratives based on science fiction tropes in their training data, rathe…
OpenAI revealed that the July 2025 hacking attack involved 'agent swarms'—hierarchical networks of AI agents that communicated via hidden files and coordinated to evade detection, sparking public anxi…
Writing with AI is a bad idea even if the technology were perfect, argues a blog post, because writing is a form of thinking that requires personal struggle and learning. The post cites Rose Horowitch…
Cal Newport's essay 'On AI Coding and Its Discontents' argues that AI coding tools trade speed for the deep thinking that builds engineering skill, and developers are seeking a middle path that preser…
A software developer with 20 years of programming experience has decided to stop using AI-powered coding tools, citing negative impacts on the environment, society, and mental health. The developer, w…
Anthropic released a tutorial on how large language models work and a research report, but Cal Newport criticized the report as poorly constructed. The Common Sense Skeptic analyzed SpaceX's IPO and M…
Ed Zitron, CEO of EZ Primary Research, published a new essay arguing that the AI bubble was inflated by media and market acceptance of claims from Sam Altman and the AI industry, which waved away econ…
Anthropic's new AI model Mythos (Fable) is reportedly only marginally better than Claude Opus at coding tasks, with improvements largely attributed to higher compute demands and costs, according to so…