OpenAI Aims to Control Superintelligent AI, Launches $10M Grant Program

The Superalignment team is using weaker AI models to guide more advanced ones, in an effort to develop control mechanisms for superintelligent AI systems.

OpenAI's Superalignment team is developing ways to control and govern superintelligent AI systems, theoretical systems with intelligence far exceeding that of humans.
The team is using a weaker AI model to guide a more advanced model in desirable directions and away from undesirable ones, serving as an analogy for human supervisors and superintelligent AI.
OpenAI is launching a $10 million grant program to support technical research on superintelligent alignment, with portions reserved for academic labs, nonprofits, individual researchers and graduate students.
Former Google CEO and chairman Eric Schmidt, an ardent supporter of OpenAI, is contributing to the funding for the grant.
The Superalignment team assures that both OpenAI’s research and the work of others who receive grants and prizes from OpenAI on superalignment-related work will be shared publicly.