Anthropic introduces the first model of the “hybrid consultation of the country”

The difference between the common model and the same consultation is the same as the two types of thinking described by the Bel-Piren-Pirenist Economist Michael Khneman in his 2011 book Reflecting fast and slowly: Quick and Wonderful-1 System 1 To think and exchange for the best things – 2 thinking.

The type of model that made Chatgt possible, known as the largest language model or llm, produces instant answers in the Prompt by asking a large neural network. These effects may be smart and curable and may fail to answer questions that need to think about steps, including simple statistics.

The llm can be forced to deliberately reform to reason if instructed to reach a strategy to follow. This strategy is always reliable, however, the models often struggle to solve problems that need a wide planning. Opena, Google, and now anthropic is now anthropic using a machine learning method that is known to get their latest models to learn to express the points receiving good answers. This requires collecting additional training information from people in resolving certain problems.

Penn says Cloude’s consultation mode has received additional information on business requests including writing and repairing the code, using computers, and to answer complex legal questions. “The things we have done to improve … Technology lessons or lessons that need a long reasoning,” Penn said. “We have from our customers than most interested in using our models of their real work.”

Anthropic says Claude 3.7 is very good in solving coding problems that require action, to determine the O1 of O1 for benches such as Bench. The company today removes a new tool, called Claude Code, designed exactly this kind of codes that help AI.

“The model is already correct in codes,” said Penn. However, “additional thinking may be good that cases may need a very complex plan – to say that you look at the Code and GOD BASE of the company.”