Nature Study: AI Agents Fail at Complex Scientific Tasks
The best AI agents achieve only half the performance of PhD experts on complex tasks. A Nature study pumps the brakes on the agent hype - while delivering some surprising nuances.
The best AI agents achieve only half the performance of PhD experts on complex tasks. A Nature study pumps the brakes on the agent hype - while delivering some surprising nuances.
The European Central Bank is reaching out to risk chiefs at eurozone banks. The topic: Anthropic's Mythos model and its ability to find vulnerabilities in financial systems.
As Anthropic and OpenAI prepare for their IPOs, public pushback against AI and data centers is growing. The timing could hardly be worse.
Investors are offering Anthropic an $800 billion valuation. Meanwhile, some OpenAI backers are starting to question their bets.
Anthropic patches a security vulnerability in the Bash tool and adds Git worktree support, plugin monitors, and faster diffs.
Anthropic launched automated routines for Claude Code alongside a redesigned desktop app. Scheduled tasks, API triggers, and an integrated terminal turn the coding agent into a round-the-clock teammate.
OpenAI's next model codenamed 'Spud' has finished pre-training. Polymarket gives it 78 percent odds of launching by end of April.
Anthropic's Long-Term Benefit Trust appointed Vas Narasimhan to the board. The Novartis chief brings pharma expertise and signals Anthropic's growing healthcare ambitions.