Useful information
Prime News delivers timely, accurate news and insights on global events, politics, business, and technology
Openai after DeepseekR1
In case you lost it in favor of The Grammy Awards last night Openai surprised the world on Sunday night with the announcement of his new “Deep research “Modality, an artificial intelligence agent available for users of the Chatgpt Pro ($ 200/month) subscription plan that are designed to save humans hours when investigating, well,” deeply “and expansively on the entire website for topics Dice and compilations of professional quality reports in specialized domains of commercial businesses.
Chatgpt Pro users (and soon, chatgpt plus, team, enterprise, and Edu) in the US Chatgpt website and applications.
Sam Altman, CEO of Operai, described the characteristic of a series of publications about his Personal account in the social network X as “as a superpower; demand experts!” He added: “It is really good and can do tasks that would take hours/days and cost hundreds of dollars.”
Deep research is based on the series of reasoning or OpenAI models, specifically taking advantage of the complete O3 model that will soon be launched (a smaller and less powerful model, O3-mini, has just been launched on Friday). The complete O3 model can analyze large amounts of information and integrate text, PDF and images into a cohesive analysis.
In Livestress posted on YouTube And available for repetition at the request, Mark Chen, head of Border Research of OpenAI, explained that “deep research is a model that conducts an investigation in multiple steps on the Internet. Discover content, synthesizes content and reasons on this content, adapting your plan as you discover more and more information. “
Chen further highlighted the importance of innovation for Openai’s vision: “This is the nucleus for our AGI roadmap. Our last aspiration is a model that can discover and discover new knowledge for yourself. “
The launch of the deep investigation marks the second in the official OpenAI agents after the launch of its cursor control and cursor control operator at the beginning of this month. And jOshua Achiam, head of alignment of the mission at the Stargate command in Operai, wrote in XBoth models can help better define the concept of an “AI agent”, a popular but foggy term in these days among companies, far beyond the company or these specific use cases.
“I feel that the term ‘agent’ wandered for the desert for a while,” Achaim wrote. “I had no basis or examples to point out. But agents such as the operator or deep research give some form to this concept. An agent is a general purpose AI that makes one or more workflows for the use of tools for you. ”
Openai’s deep research achieves a new and higher score at the last reference point of ‘Humanity’s last exams
Deep research has established new reference points for precision and reasoning.
Isa Fulford, a member of the OpenAI research team, shared in the live transmission of YouTube that the model achieves “a new maximum of 26.6% precision” in “” “The last examination of humanity”A relatively new reference point designed to be the most difficult for any model of AI (or human, for the case) to complete, covering 3,000 questions on 100 different topics, such as translating old inscriptions in archaeological findings.
In addition, its ability to navigate the web, dynamically reason, and cite Fuentes distinguishes it precisely from the previous AI tools.
“The model was trained using end -to -end reinforcement learning in navigation and reasoning tasks,” Fulford said. “He learned to plan and execute several steps trajectories, reacting to real -time information and backing when necessary.”
An outstanding feature of deep research is its ability to handle tasks that would otherwise take hours of humans or even days.
During the announcement, Chen explained that “deep research generates results that resemble comprehensive and fully cited research work, something that an analyst or expert in the field could produce.”
Applications and use cases
Use cases for deep research are as diverse as shocking.
The official Operai account in X declared that it was “built for people who work in intensive knowledge in areas such as finance, science, politics and engineering and need thorough and reliable research.”
It also seems valuable for consumers looking for personalized recommendations or conducting detailed products, according to the examples shared by OpenAI in its official publication blog publicationwhich includes a detailed research evaluation of the best snowboard for someone to buy.
Altman summarized the versatility of the tool, writing: “Try your hardest work task that can be resolved just by using the Internet and see what happens.”
A personal medical success story of deep research
Felipe Millon, leader of the OpenAI government market, shared a deeply personal description of how deep research hit his family. Write in a series of publications in XHe described his wife’s battle with bilateral breast cancer and how the AI tool became an unexpected ally.
“At the end of October, my wife was diagnosed with bilateral breast cancer. During the night, our world put on top legs, ”Millon wrote.
After a double mastectomy and chemotherapy, the couple faced a critical decision: whether or not to continue radiotherapy. The situation was full of uncertainty, since even its specialists provided mixed recommendations. “For its specific case, it is completely in a gray area,” Millon explained. “We feel stuck.”
Having a preview of deep investigation, Millon decided to raise his wife’s surgical pathology and ask if the radiation would be beneficial. “What happened later was amazing,” he wrote. “Not only did he confirm what our oncologists mentioned, it was deeper. He cited studies that he had never heard and adapted when we added details such as his age and genetic factors. “
The specific notice he used was:
“Read the Surgical Pathology (attached) report that contains information about bilateral breast cancer. Then, investigate whether radiation would be indicated for this patient after 6 rounds of TCHP chemotherapy, according to the type of breast cancer. I want to understand the pros and cons for this patient, how likely it would be to reduce the chances of recurrence and if the benefits exceed the potential long -term risks. “
Millon and his wife verified each study cited by the model, finding that they are precise and very relevant. “Soon we see another specialist, but we are already more sure of our decision,” he wrote. “It gave us peace of mind when we needed it most.”
Availability and what follows?
Currently, deep research is available to Chatgpt professional users, with plans to expand to PLUS and TEAM levels, followed by business and educational markets.
As Chen warned: “It is still possible that it is hallted, so when you are making reports, be sure to verify the sources yourself.”
The ability of the model to think autonomously for prolonged periods also makes it intensive in resources, and OpenAi is currently working to optimize its performance for broader accessibility.
Operai has also suggested future integrations with personalized data sets, which would allow organizations to take advantage of the patented research tool.
For Millon, the impact of deep research is already clear. “We often speak internally in Operai about the moments when you feel the AGI ‘, and this was one of them,” he wrote. “This will change the world.”
