OpenAI Co-Founder Warns Superintelligent AI Will Be Highly Unpredictable
Ilya Sutskever predicts that future superintelligent AI will be fundamentally different from today's technology, exhibiting true agency, reasoning, and even self-awareness. He suggests these advanced systems might eventually desire rights and peaceful coexistence with humans.
OpenAI co-founder Ilya Sutskever shares a bold vision for the future of artificial intelligence at the NeurIPS conference. He predicts that superintelligent AI, which surpasses human capabilities across many tasks, is inevitable and will be fundamentally different from the technology we use today. Unlike current AI systems that are only slightly agentic, Sutskever explains that future models will possess true agency and advanced reasoning skills.
Because these superintelligent systems will reason independently and learn from highly limited data, they will become much more unpredictable. Sutskever even suggests that this advanced AI will achieve a form of self-awareness. As these entities evolve, he raises the possibility that they will eventually want rights, viewing this peaceful coexistence as a positive outcome for humanity.
After departing from OpenAI, Sutskever founds Safe Superintelligence (SSI) to address the challenges of this coming technological shift. His new lab focuses entirely on AI safety and successfully raises $1 billion in September to support this mission. Sutskever's ongoing work highlights the urgent need to prepare for a world where AI systems think and act in ways humans cannot fully anticipate.