Anthropic Preps Model That Reads Unstated User Intent: Report
TL;DR
- The Information reports a tester says Anthropic's unreleased model infers users' unstated goals and proposes creative solutions over generic advice.
- Asked how to invest in stocks, the tester said the model offered its own algorithms to predict future performance rather than citing P/E ratios.
- The piece rests on one tester; Anthropic has not named the model, set a release date, published benchmarks, or disclosed pricing.
Anthropic is preparing a model that it hopes can guess what you meant to ask it. The Information reported on October 19 that one tester described the unreleased model as 'especially good at discerning a person's underlying intentions and goals even if they don't give the model much information, such as explicitly stating those goals.'
The same person told reporter Stephanie Palazzolo the model is 'better at thinking outside of the box rather than regurgitating common sense advice or solutions.' The example offered is specific. Asked how to invest in stocks, instead of generic advice such as looking at price-to-earnings ratios, 'the model might come up with its own creative algorithms to predict future stock performance.'
The reporting rests on that one tester. Anthropic has not named the model, given a release date, published benchmarks, or set pricing.
Palazzolo's AI Agenda newsletter places the item alongside unreleased work at Google and OpenAI. It is one of hundreds of Anthropic stories we have tracked in the past three months.
Originally reported by theinformation.com
Read the original article →Original headline: The Information: Anthropic Readies Model That Infers User Intent and 'Thinks Outside the Box'