A new “in -depth research” tool for Chatgpt. “It will be at the level of a human analyst”

The American Openai company presented a new feature for the Chatgpt conversational tool on Monday. The new instrument is able to achieve in a few tens of minutes what a man would take several hours.

A new feature for Chatgpt Photo: Shutterstock

Openai, who face a competition from the Chinese platform Deepseek, introduced “Deep Research” a new “in -depth research” tool for Chatgpt, AFP, quoted by Agerpres.

The announcement was made in the context in which the appearance of the Chinese start-up, which proposes a powerful and low cost conversational robot, questions the economic model in this sector.

Deep Research is the new tool that can work for your benefit independently: you give it an instruction (prompt), and
Chatgpt will find, analyze and synthesize hundreds of online sources to create a comprehensive report, on the level of an analyst (human)“, the company assured.

It is a system that – this is just an estimate on my part (…) – can reach a single -digit percentage (between 1 and almost 10%, no) of all the valuable economic tasks in the world“, said Sam Altman, the Openai CEO, at the Tokyo Business Forum.

Sam Altman met on Monday with the SoftBank CEO, Masayoshi Son, at a forum that brings together about 500 Japanese companies, where SoftBank and Openai have announced a joint venture to bring you advanced to the archipelago companies.

Deepseek puts pressure on Openai

Openai and the Japanese investment giant in Technology SoftBank are partners in “Stargate”, a new project with investments of at least $ 500 billion in AI infrastructure, recently revealed by US President Donald Trump, notes AFP.

This project is included in the growing Chino-American rivalry. The Deepseek chatbot amazed Silicon Valley by his ability to equal his American rivals at a fraction of their cost, working with lower-power chips and consuming less energy.

Last week, Openai accused Chinese companies of trying to copy its models using a process called distillation-transferring knowledge from a large model, already trained, to a smaller one-contrary to the conditions of use of the Californian start-up.

“We have no intention to follow Deepseek right now. We will simply continue to build excellent products and be world leaders in terms of modeling capacities. That will work well, “said Sam Altman, Monday, in front of the press.

In an interview with the Japanese daily Nikkei, the Openai CEO had estimated that China recently reaches “significantly” the technologies you develop in the United States. He considered Deepseek as “a good model”, which highlights competition Serious for “reasoning” technology, but “its capacity level is not new,” according to statements made for the Japanese daily.

If authoritarian regimes use a strong you have a strong to strengthen their power, it would be a bad thing“, Altman also warned.