OpenAI
OpenAI’s new policy paper proposes a “public wealth fund” that hands every citizen a stake in AI-driven growth – but critics argue it just ties the public’s wellbeing to the tech industry’s boom-and-bust cycles instead of expanding tested safety nets.
Researchers warn that AI capabilities are increasingly discovered rather than designed, and that the window to understand how these black-box systems reason is closing as their influence grows.
Anthropic has made Fable 5, the first commercial model from its advanced Mythos class, available to the public — while reserving the unrestricted Claude Mythos 5 for vetted cybersecurity partners.
Cisco has launched Cloud Control, a toolset that lets companies build their own armies of defender AI agents, a response to a world where cyberattacks increasingly come from swarms of hostile agents.
With AI bills running far ahead of budgets, companies are turning to model routing, sending easy tasks to cheap models, in a shift that could squeeze the economics of OpenAI and Anthropic.
SoftBank’s Masayoshi Son says OpenAI’s next model is being designed by another model, a sign, he argues, that superintelligence is just two years away.
OpenAI says one of its general-purpose reasoning models has disproved an 80-year-old conjecture by mathematician Paul Erdős, a result validated by mathematicians but one that still leaves the deeper problem unsolved.
Three solo founders — behind Daymaker, OpenClaw and Base44 — have turned Sam Altman’s ‘one-person company’ idea into reality, showing how AI tools now let individuals build and sell businesses at startling speed.
Sixteen researchers behind the Leiden Declaration, endorsed by the International Mathematical Union, warn that AI and tech-industry money threaten the values, autonomy and future of mathematical research.
A Walmart-backed startup, Eko, is hand-building an “AI-ready” product catalogue — photographing millions of items from every angle — so AI shopping agents like ChatGPT and Gemini can describe and sell them accurately.
Multiverse Computing ran a pretrained Llama model through quantum circuit blocks on IBM’s 156-qubit machine and cut its perplexity while adding almost no parameters — the first end-to-end “quantum enhancement” of a production LLM.
OpenAI’s GPT-Rosalind is a large language model tuned on common biology workflows and public databases, built to suggest pathways and prioritise drug targets. Citing bioweapon risk, OpenAI is limiting access to vetted US-based researchers for now.