OpenAI Co-founder Predicts Superintelligent AI and Its Implications
At the NeurIPS AI conference, OpenAI co-founder Ilya Sutskever shared his insights on the future of artificial intelligence, predicting the emergence of "superintelligent AI" that surpasses human capabilities in many tasks. He believes this advancement will be qualitatively different from current AI, exhibiting characteristics not yet seen.
Sutskever highlighted key distinctions of superintelligent AI: true agency, unlike today's "slightly agentic" systems; enhanced reasoning abilities leading to increased unpredictability; comprehension from limited data; and potential self-awareness. xAI's recent upgrades could be seen as a step in this direction.
Intriguingly, Sutskever also touched upon the possibility of AI seeking rights, suggesting a future where AI desires co-existence and recognition. This raises ethical questions explored by articles like this one on AI security and ethics.
Following his departure from OpenAI, Sutskever founded Safe Superintelligence (SSI), a lab dedicated to ensuring the safety of general AI. SSI recently secured substantial funding, reflecting the growing concern and investment in this area, similar to Databricks' recent funding round.