Anthropic has published new research that sheds light on how its AI assistant Claude “thinks” internally. Two recent papers explore the model’s internal mechanisms through a novel interpretability approach the company compares to an “AI microscope.” This research reveals several surprising findings about Claude’s cognitive processes, including how it handles multiple languages, plans ahead when …
Analyst firm Gartner forecasts that global spending on generative AI will reach $644 billion in 2025, representing a 76.4% increase from 2024. According to the report detailed by Sean Michael Kerner in VentureBeat, hardware will account for approximately 80% of this spending, with devices alone claiming $398.3 billion. John Lovelock, distinguished VP analyst at Gartner, …
A new study by the AI Disclosures Project claims OpenAI likely trained its GPT-4o model on paywalled books from O’Reilly Media without a licensing agreement. Researchers Tim O’Reilly, Ilan Strauss, and Sruly Rosenblat analyzed how well different OpenAI models recognized content from O’Reilly books. As reported by Kyle Wiggers for TechCrunch, the team found that …
Midjourney has released V7 Alpha, its first new AI image generation model in nearly a year, featuring voice prompting capabilities and a faster draft mode. The launch comes a week after OpenAI debuted a new image generator in ChatGPT that quickly gained popularity. According to Midjourney CEO David Holz, V7 represents a “totally different architecture” …
OpenAI has reversed its decision to cancel the consumer release of its o3 reasoning model. CEO Sam Altman announced on X (formerly Twitter) that both o3 and a next-generation successor, o4-mini, will launch in “a couple of weeks.” According to Kyle Wiggers reporting for TechCrunch, this change relates to OpenAI’s upcoming GPT-5, which will incorporate …
Microsoft is rolling out its Copilot Vision feature to Windows and mobile platforms, extending the AI assistant’s capabilities beyond web browsers. Previously limited to Edge webpages, Copilot Vision can now analyze real-time video from mobile cameras and interact with content across Windows applications. The expanded vision capabilities allow Copilot to provide practical assistance such as …
Oumi has released HallOumi, an open-source claim verification model designed to detect and prevent AI hallucinations. According to Sean Michael Kerner of VentureBeat, the tool analyzes AI-generated content on a sentence-by-sentence basis, providing confidence scores, specific citations, and human-readable explanations. Led by ex-Apple and Google engineers, Oumi aims to build an unconditionally open-source AI platform. …
Researchers from Sentient, the University of Washington, Princeton University, and UC Berkeley have introduced Open Deep Search (ODS), a new open-source framework designed to match the capabilities of proprietary AI search solutions. The system combines reasoning agents with web search tools to enhance the performance of large language models. According to the research team led …
Google has added a new “Discover sources” feature to NotebookLM that searches the web for relevant content. As reported by Abner Li, this update allows users to find information without manually adding files and links. After clicking the new button in the Sources pane and entering a topic of interest, NotebookLM delivers up to 10 …
Adobe has released version 25.2 of Premiere Pro, featuring new AI-powered tools including Generative Extend. This feature, powered by Adobe’s Firefly AI model, allows editors to extend video clips by up to two seconds and ambient audio by up to ten seconds. As Jess Weatherbed reports, the tool is free for a limited time before …