A preprint posted to arXiv on 29 July 2026 and revised on 7 August 2026 introduces 'shadow evaluations,' a method for measuring AI agents' ability to conduct open-ended research, and reports that fron…
A new study led by Peter Kirgis and Sayash Kapoor at Princeton University found that AI agents are not yet capable of conducting open-ended AI research, as they lack the judgment and creativity needed…
A study posted on arXiv in late July found that AI systems are not yet capable of fully automating open-ended research, with an AI system built by Princeton University researchers scoring 2/6 and 1/6 …
A new analysis by Iain on the blog argues that the divide between AI 'foomers' and 'doomers' in software development is a systems failure, not a cultural one, citing NBER Working Paper No. 35275 which…
The UK's liver allocation algorithm appears to discriminate by age, with some younger patients seemingly unable to receive a transplant regardless of illness severity, according to researchers Arvind …
Princeton University professor Arvind Narayanan, in his keynote at the International Conference on Machine Learning in Seoul, argued that AI will not suddenly eliminate all jobs, but that future work …
Tim O'Reilly argues that the current AI industry's focus on frontier models and heroic inventors is misguided, drawing on political scientist Jeff Ding's diffusion theory which holds that economic lea…
Simon Willison explores the importance of data provenance and accountability in AI systems, arguing that the ability to trace outputs back to source columns is more critical than the spectacle of AI c…
AI engineers face a higher risk of job displacement from artificial intelligence than other developers, according to an engineer who argues that the broad and vague definition of the role makes it vul…
Google claimed at its developer conference that a team of AI agents built an entire operating system from a single prompt for $916 in API fees. Independent researchers have raised concerns that the cl…