Claude 3.7 Sonnet: one more step towards a human reasoning

Anthropic has just deployed its new model, called Claude 3.7 Sonnet. This one seeks to adopt a mode of hybrid reasoning, as close as possible to the reflection that a human could have. In other words, this artificial intelligence is capable of quickly answering simple questions, but also of providing further reflection if the context invites it.

A hybrid AI model to reason like humans

This new model is trying to simplify the daily life of users; Rather than selecting between a so -called “rapid” model and a “complex” Claude 3.7 reasoning model independently selects the level of iterations required in its reflection. This artificial intelligence is therefore able to answer a simple question very quickly, such as spending more time on more advanced issues.

This “hybrid” model feature is not completely new, Openai would work on a similar system for the next version of its model, GPT-5.

New features for Claude 3.7 SONNET

Anthropic also supports the release of Claude 3.7 Sonnet of a brand new brand new tool: Claude Code, designed to help developers on computer projects. This new model also has a new feature that takes the form of a notebook. Integrated into the interface, the latter displays the ideas in real time by which Claude makes his reflection.

Following as the model is reasoning, the user can refer the AI ​​in a better direction by refining his prompt. A “break” button is also provided in this sense in the interface of Claude 3.7.