Claude Opus 4.8 is being released with a focus on honesty. Models are trained to avoid claims they cannot support, and to reduce confident conclusions drawn from thin evidence. Early testers report that Opus 4.8 flags uncertainties more often and makes fewer unsupported claims. In evaluations, it is about four times less likely than its predecessor to allow flaws in code that it writes to pass unnoticed. Users can also control how much effort the model applies to a task, with higher-effort responses using more tokens. Anthropic is additionally launching dynamic workflows in research preview, enabling Claude to plan work, run hundreds of parallel subagents in one session, and verify outputs before responding.
"Anthropic is releasing Claude Opus 4.8 on Thursday, and the company is touting the model's "honesty." According to Anthropic, it trains "all [its] models to be honest - for instance, to avoid making claims that they can't support." But it notes that "a general problem with AI models is that they sometimes jump to conclusions, confidently presenting their work as making progress despite thin evidence.""
"The AI lab claims that early testers have found that 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 "around 4x less likely than its predecessor to allow flaws in code it's written to pass unremarked.""
"In addition to the honesty improvements, with Opus 4.8, users can direct the amount of effort Claude puts into a task. Higher-effort responses will use more tokens, giving users the option of lower-effort responses if they don't want to burn through their rate limits as quickly."
"Anthropic is also launching a feature called "dynamic workflows" in research preview, which the company says will let Claude "take on even bigger tasks." With dynamic workflows, "Claude can plan the work and then run hundreds of parallel subagents in a single session (and with Opus 4.8, the agents can run for even longer). It then verifies its outputs before reporting back to the user.""
Read at The Verge
Unable to calculate read time
Collection
[
|
...
]