New Claude, New Realities

Anthropic dropped Opus 4.8 on Thursday, raised $65 billion at a $900 billion valuation, and announced a feature that lets Claude spawn hundreds of parallel coding agents to tackle multi-day engineering projects. The same day, Blue Origin's New Glenn rocket exploded during testing, Amazon killed its internal AI usage leaderboard after employees gamed it, and a startup called Shift offered free house cleaning in exchange for recording cleaners to train robots. Here is what happened and why it matters.

Anthropic's Opus 4.8 and the Honesty Push

Anthropic released Claude Opus 4.8 on Thursday, and the company is emphasizing the model's honesty. According to Anthropic, it trains all its models to avoid making claims they cannot support, but it notes that "a general problem with AI models is that they sometimes jump to conclusions, confidently presenting the work as making progress despite thin evidence." The lab claims early testers found Opus 4.8 is "more likely to flag uncertainties about its work and less likely to make unsupported claims." In the company's evaluations, Opus 4.8 is roughly 4X less likely than its predecessor to let flaws in code it wrote pass unremarked.