There are not two ways, there is a new sense of urgency in Openai. Two days after having released O3-Mini around the world, the company made a surprise announcement Sunday evening, RIn -depth research. The new feature allows Chatgpt to find, analyze and synthesize hundreds of websites and online sources to create reports “at the level of a research analyst”.
In addition to the usual text questions, users can download files, including PDFs and calculation sheets, when they encourage Chatgpt in this way. The chatbot will then put “between 5 and 30 minutes” to compile an answer, a side panel documenting the agent’s progress and quotes as and when. “He accomplishes in tens of minutes which would take a human of many hours,” explains Openai about the new feature.
“Our ultimate aspiration is a model that can discover and discover new knowledge for itself,” said Mark Chen, research director in Openai, during the revelation of the company. “It is at the heart of our [artificial general intelligence] Roadmap. “”
Regarding the limits, Openai says that the Chatppt can sometimes hallucinate facts or make incorrect inferences during in -depth research, although “at a significantly lower pace” than the other current models. In addition, the agent may sometimes have trouble differentiating between authority and rumors. Users can also notice certain formatting errors. “We expect all these problems to improve quickly with more use and time,” notes the company.
If all of this seems familiar, it is because Google’s advanced suite understands its own in -depth research functionality, which does not only share the same name, but largely offers the same set of capacities. A significant difference between the two is that Google offers access to Gemini Advanced through its $ 20 per month A Premium AI plan. On the other hand, you will need a $ 200 chatgpt pro plan to start using the OPENAI version of Deep Research today.
“In -depth research in Chatgpt is currently very intensive,” explains the company, adding that it will limit pro users to 100 queries per month. “The more time it takes to search for a request, the more an inference calculation is needed.”
OPENAI says he is working on a deep research version powered by a smaller and more profitable model. In turn, this will allow the company to offer “much higher rate limits”. In the meantime, Openai hopes to get the tool in the hands of users more “in about a month”, after a series of security tests. As for most other recent versions of the company, European users will have to wait before they can try the tool for themselves, with in-depth research not yet available for people in the United Kingdom, Switzerland and in the broader European economic zone.
This article originally appeared on engadget to
