Engineering
Towards Full Pipeline FP8 Reinforcement Learning for LLMs
The paper identifies that FP8 quantization noise in RL pipelines distorts importance ratios, causing training instability. It proposes ‘Calibrated Clipping,’…
News
Anthropic Releases Opus 5.5 With Lower Prices and Fable-Level Performance
Anthropic released Claude Opus 5.5, describing it as the strongest-performing model the company has tested to date. The model matches…
News
Nvidia Commits $2 Billion to Brookfield AI Infrastructure Fund
Nvidia has committed $2 billion as an anchor investor in Brookfield Asset Management’s new artificial intelligence infrastructure fund, alongside the…
News
FTC Chair Suspicious of Calls for AI Antitrust Exemptions
FTC Chairman Andrew Ferguson said “everyone should be deeply suspicious” of AI companies that seek antitrust exemptions while simultaneously lobbying…
News
Will AI Chip Export Controls Make a Difference Against China?
Members of Congress are pushing to tighten restrictions on AI chip exports to China through the National Defense Authorization Act,…
News
Anthropic’s Annualized Revenue to Top $100 Billion in 2026, NYT Says
Anthropic’s annualized revenue is on pace to exceed $100 billion in 2026, according to a New York Times report citing…
Opinions
A Labor Day Prescription for Preserving Fair Employment in the Digital Age
Charlotte A. Burrows, former chair of the U.S. Equal Employment Opportunity Commission, argues that AI hiring tools can reject qualified…
Opinions
When AI does science, who is accountable for mistakes?
This Nature commentary examines how responsibility for scientific errors should be assigned as AI systems take on increasingly autonomous roles…
Opinions
Why AI resembles a charismatic religious leader
Religious studies scholar Morgan Shipley argues that AI chatbots function similarly to charismatic religious leaders by producing personalized, seemingly authoritative…
Opinions
AI is advancing fast. Women are being left behind in the workforce and leadership.
A UN Women report launched at UNGA81 finds that women hold only around 22% of AI professional jobs globally and…
Engineering
The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks
Researchers introduce Taste-Bench, a benchmark that measures an LLM agent’s “taste” — its ability to make good long-horizon decisions at…
Engineering
Realtime-Venus: A full-duplex interaction system with asynchronous delegation
Realtime-Venus is a full-duplex conversational AI system built from two separately trained 9B models — one for audio-visual interaction and…