Isla Fulford, a researcher at OpenaiThere is a premonition Profound research It will be a hit even before it is released.
Fulford has helped build Artificial intelligence The agent, automatically discover the web, decide which links to click, read what and what needs to compare into an in -depth report. Openai for the first time conducted in -depth research within the internal; Whenever it went down, Fulford said, she was submerged in queries from her colleagues eager to return. The number of people who are making me makes us quite excited, Mr. Fulford said.
Since broadcasting to the public on February 2, Deep Research has also been shown to be a hit for many users outside the company.
Today, Deep Deep Research wrote 6 reports Posted on x A few days after the product is released. This is really great. Congratulations to the people behind it.
Deep research is the product that AI really has a significant part of the policy making community in DC to start feeling AGI, written Dean Ball, a colleague at George Mason University, who specializes in AI policy.
The in -depth study is available as part of Chatgpt for The plan, costs $ 200 per month. It needs a query, such as writing a letter to Massachusetts Health Insurance, or tell me about the scope of Wired insurance about Efficiency of the Ministry of the GovernmentAnd then make a plan, search for relevant websites, through their content and decide which links to click and which information deserves further investigation. After discovering sometimes dozens of minutes, it summarizes its findings into a detailed report, which may include quotes, data and charts.
Many tools are currently considered AI agents, which are basically chatbot connected to simple programs without much sophistication. The deep research model itself has gone through an artificial theory before giving a plan and moving forward with each step. The model provides details of this reason behind its research in an extra window.
Sometimes, it is like 'I need to go back, this does not seem to have that prospect,' 'Josh Tobin, another Openai researcher involved in building a deep research. It is great to read some of those orbit, just to understand how the model is thinking.
Openai clearly considers deep research as a tool that can take more office work. This is a thing that we can expand the scale, Mr. Tob Tobin said, adding that agents can be trained to complete the specific white neck work. For example, an agent with access to internal data of the company can prepare or present quickly. Tobin says that the longer goal is to build an agent that is not only good at building reports through searching on the web, but also good in many other types of tasks.
Because in -depth research was trained to analyze and summarize human writing, Tobin said that his group was surprised to see many people using it to create codes. This is an interesting topic to pull, he said. We are not completely sure what to do for it.