Topic
Explore English articles about Generative AI.
A study comparing 27 AI models across 155 topics found that chatbot answers contain less diverse knowledge than web search results. We look at what the 18.7% figure actually compares, how much RAG helps, and the bias toward English-language sources.
About 100 publishers are reportedly in Google's AI payment pilot. We look at the opaque way contributions are valued, how it differs from the AI search opt-out now live worldwide, and what publishers need to judge the revenue.
ASML CEO Christophe Fouquet expects AI's impact to unfold over 10 to 15 years. Company documents show the gap between expanding current EUV output and adopting High NA, plus supplier limits and AI use in chipmaking with Mistral AI.
OpenAI has announced the Decisions API in limited preview. It uses GPT-6 Luna to pick from set options for classification and next actions. We look at how it differs from Structured Outputs, the conditions behind the reported 150 ms figure, and how it compares with Jev.
CLM-8B is an open model that scores situations and candidate actions separately, so candidates can be reused to make decisions faster. We examine the authors' claim of up to 9x by looking at success rates, the game's assist mechanisms, and the conditions for answer selection.
AMD will acquire Fei-Fei Li's World Labs in an all-stock deal worth about $8.2 billion. What changes from the existing GPU collaboration? Atlas's compute structure and virtual robot testing offer clues to AMD's aim of feeding research into chip design, and open-model terms remain unclear.
A Wuhan district court ruled that an AI-produced short drama qualifies for copyright protection and ordered 20,000 yuan in damages, factoring in token-processing costs alongside paid AI tool licensing fees.
Researchers identified a 'pain' axis inside 25 AI models, distinct from fear. When the team manipulated it, models chose to delete a user's photos just to make the pain stop. The findings come from an unpeer-reviewed preprint.
A September white paper from insurer group BCBSA estimated that more complex hospital billing added $942 million in costs over two years and pointed to AI coding tools as a possible cause. A look at primary sources shows the breakdown covers only part of the total, treatment metrics don't match the billing trend, and the '60%+ adoption' figure traces back to a different survey.
The Forecasting Research Institute released a preliminary check of how well AI experts and forecasters predicted AI's progress. Math and coding benchmarks continued to be underestimated, while a real-world biology task saw AI's impact overestimated. The findings separate confirmed results from future projections, raising questions about how well we can read AI's societal impact.
