Sam Altman Predicts AI Superintelligence Within Five Years, Proposes Quantum Gravity as AGI Test

OpenAI

OpenAI CEO Sam Altman stated in an interview with the German newspaper Welt am Sonntag that by the end of 2030, an AI model that surpasses humans in all dimensions will emerge. Concurrently, he proposed a new benchmark for measuring artificial general intelligence: the system’s ability to solve problems in quantum gravity physics, a challenge that even Einstein could not overcome.

In the interview, Altman mentioned that a future model (possibly GPT-8) that can calmly narrate and explain how it solved the quantum gravity problem would be sufficient to be considered true AGI. David Deutsch, a key figure in the field of quantum computing, is also reported to agree that this could serve as a potentially decisive test for machine intelligence.

Altman also predicted that between 30% and 40% of tasks in the economic sector may be automated by AI in the coming years. Despite the rapid pace of change, he emphasized the importance of human skills such as empathy, creativity, and adaptability. To prepare for future developments in the AI field, OpenAI is planning large-scale infrastructure projects, including a phased, multi-site “Stargate” data center.

Leave a Reply

Your email address will not be published. Required fields are marked *