Openai has advanced artificial intelligence development by introducing a new tool that claims to produce reports comparable to those of research analysts.
Developers of Chatgpt have dubbed the tool “Deep Research,” stating that it can accomplish tasks that would take humans hours in just 10 minutes.
This announcement comes shortly after a San Francisco-based company accelerated its product release in response to the progress made by Openai’s competitor, Deepseek.
“Deep Research” is an AI agent that allows users to delegate tasks and is powered by Openai’s latest cutting-edge model, O3 version.
Openai explained that deep research scours hundreds of online sources, analyzes, integrates, creates comprehensive reports, and sifts through massive amounts of text, images, and PDFs.
The company views tools like the Chatgpt button as essential steps towards achieving artificial general intelligence, a concept that aims to match or exceed human intelligence in various tasks.
Last month, Openai unveiled an AI agent named Operator, claiming it can manage an online store based on photos of shopping lists, albeit only in the US preview version.
In a demonstration video released on Sunday, Openai showcased Deep Research analyzing the translation app market, stating that each task takes 5-30 minutes to complete with proper sourcing.
Openai highlighted that deep research targets experts in fields like finance, science, and engineering but can also be utilized for car and furniture purchases.
Leveraging Openai’s latest “reasoning” model, O3, deep research processes queries slower than traditional models and has a partially disclosed entity named O3-mini, a slimmed-down version of O3.
The full capabilities of the O3 model were outlined in the recent international AI safety report, prompting concerns from experts like Yoshua Bengio about the potential risks posed by AI advancements.
After the newsletter promotion
Deep surveys are accessible to Openai’s protia users in the US for $200 (£162 per month), with a monthly limit on queries due to processing constraints. Not available in the UK and Europe.
Andrew Rogoyski, director of an AI Research Institute affiliated with Sally University, cautioned about the potential dangers of blindly relying on deep search tools without conducting thorough verifications of their outputs.
“Knowledge-intensive AI faces a fundamental challenge. Human validation and verification are crucial to ensure the accuracy of machine analysis,” said Rogoyski.
Source: www.theguardian.com
