Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

OpenAI Holds Surprise Livestream to Announce Multi-Step 'Deep Research' Capability - Slashdot

Feb 03, 2025 - slashdot.org
OpenAI recently announced a new feature called "Deep Research" during a livestream event from Tokyo. This capability, initially available to ChatGPT Pro users, is designed to perform multi-step research by discovering, synthesizing, and reasoning about content. It also asks clarifying questions to ensure accuracy and relevance. The feature will soon be integrated into other OpenAI products, with a dedicated page on their website describing it as an agent that uses reasoning to handle large amounts of online information.

The announcement has sparked discussions on X.com, with some users drawing parallels to OpenAI's previous project, code-named "Strawberry." This project, reported by Reuters in July 2024, aimed to enhance AI's human-like reasoning skills, enabling autonomous and reliable internet navigation for deep research tasks. Despite the secrecy surrounding Strawberry, it is believed to significantly improve AI reasoning capabilities, a crucial step towards achieving human or super-human-level intelligence. OpenAI's CEO, Sam Altman, has emphasized the importance of advancing AI's reasoning abilities.

Key takeaways:

  • OpenAI announced a new capability called Deep Research, which is a model designed for multi-step research tasks, available first to ChatGPT Pro users.
  • Deep Research is described as an agent that synthesizes large amounts of online information and asks clarifying questions to ensure accurate research.
  • The concept of "deep research" was previously associated with OpenAI's internal project code-named "Strawberry," which focuses on human-like reasoning skills.
  • OpenAI aims to improve AI models' reasoning capabilities, with the belief that reasoning is crucial for achieving human or super-human-level intelligence.
View Full Article

Comments (0)

Be the first to comment!