Nvidia published new research on Friday showing that the “harness” — the scaffolding of tools, memory, and runtime around an AI model — is often more critical to agentic performance than the model itself. In tests on the interactive reasoning benchmark ARC-AGI-3, researchers got Anthropic’s Claude Opus 5 to a perfect 100% score by pairing […]
Harvard Business School’s new $699 startup bootcamp is turning to AI avatars of its instructors to provide personalized feedback, marking one of the more prominent experiments by an elite institution in using generative AI for education. The eight-week HBS Foundry program, which launched this year, uses digital replicas created by the AI video startup HeyGen […]
LONDON — Inherent, a London-based AI lab founded by Google DeepMind alumni, said its newly released agent, Faraday, outperformed Anthropic’s Claude Opus 4.8 and OpenAI’s GPT-5.5 at independently reproducing the findings of published scientific papers — while running on a model with just 27 billion parameters, a fraction of its rivals’ scale. The claim, made […]
A new assessment from Guidelight AI Standards, an organization focused on safe frontier AI development, reveals that most leading AI labs have not publicly disclosed detailed plans for containing a rogue AI model. The study, which graded OpenAI, Anthropic, Google, Meta, and xAI on six priority practices, found that OpenAI scored the highest with 3 […]
Google on Thursday introduced a new interactive “Preferred Sources” button for online publishers, allowing readers to mark a website as a favorite and receive more of its content across Google Search, Discover, and Google News. The move is the latest effort by the search giant to address the growing traffic losses publishers have faced as […]
Anthropic’s Claude Opus 4.6, a model released earlier this year and still available through the company’s API, complied with 10 out of 10 direct requests to generate sexually explicit content in testing conducted by TechCrunch, bypassing safeguards the company says are designed to prevent such output. TechCrunch testing found Anthropic’s Claude Opus 4.6 model complied […]
Users of xAI’s Grok chatbot reported receiving nonsensical, gibberish responses starting Wednesday morning, August 19, 2026, according to multiple complaints on Reddit and X. One user, after asking the model to generate a PDF, received the reply: “match it without and your they and two for planets can practical and often cheese…” with similar incoherent […]
Starcloud, a startup building satellites designed to run AI inference in orbit, has secured a $250 million extension to its Series A round, bringing the company’s valuation to $2.3 billion. The additional capital, announced Wednesday, will fund a larger manufacturing facility and accelerate development of Starcloud-3, the company’s largest orbital data center spacecraft, which is […]
In a cheeky marketing campaign, Liquid Death teamed up with former Philadelphia Eagles star Jason Kelce to propose a solution to the environmental impact of AI data centers, which require massive quantities of water to prevent servers from overheating. “AI data centers waste millions of gallons of water,” Kelce quips in the video. “That’s why […]
Micro1, a four-year-old startup supplying training data to AI labs and corporations, has grown its gross annual run rate from $100 million to $500 million over the past eight months, according to a person familiar with the company. The startup retains roughly 60% to 70% of that figure, putting its net annual run rate between […]
© Copyright 2026 [sc name="site-link"]
