Archive · 3 Jun 2025
OpenAI's Altman: AI to Discover New Knowledge Soon
OpenAI's CEO, Sam Altman, thinks AI is getting smarter.
He believes AI agents will soon be able to help us learn new things and solve tough problems, not just help with simple tasks.
It's like AI is becoming a teammate, similar to a junior employee.
Altman thinks this could happen as early as next year, 2026.
Interestingly, a former OpenAI scientist, Ilya Sutskever, even wanted to build a special shelter before they released very advanced AI, as detailed in a new book about the company's history.
Sam Altman believes AI is evolving from a tool to a team member.
Altman predicts AI agents could discover new knowledge by 2026.
AI agents are currently being used like junior employees.
Ilya Sutskever proposed building a 'doomsday bunker' before releasing AGI.
These revelations come from an upcoming book about OpenAI's internal events.
- Who
- OpenAI, led by Sam Altman, with a past proposal from Ilya Sutskever.
- What
- AI agents may soon discover new knowledge and solve complex problems.
- When
- AI agents could achieve new breakthroughs by 2026.
- Where
- Mentioned at the Snowflake Summit 2025 and within OpenAI's discussions.
- Why
- To advance AI capabilities beyond basic assistance.
- Company
- OpenAI
- Focus
- AI Agents
- Altman's Prediction
- Discovering new knowledge and solving business problems.
- Anticipated Timeline
- 2026
- Sutskever's Proposal
- Building a protective bunker before releasing AGI
