OpenAI says it is launching a program to award ten $100,000 grants to fund experiments to create a democratic process for deciding what rules artificial intelligence systems should follow — “within limits defined by law.”
The grant program was launched after OpenAI called for an international AI regulatory body to regulate nuclear power.
The founders of OpenAI Sam Altman, Greg Brockman and Ilya Sutskever argued in their proposal to create such an agency that the pace of innovation in artificial intelligence is so fast that we cannot expect the current authorities to adequately curb the technology, an announcement today. also beats
Specifically, OpenAI says it aims to fund individuals, groups and organizations to develop evidence for a “democratic process” that could answer questions about the limits of artificial intelligence. The company wants to learn from these experiments and use them as a basis for a more global and ambitious process.
“While these preliminary experiments are not intended to be decision-making (at least for now), we hope they will explore issues related to the decision and create new democratic tools that can more directly inform decisions in the future,” the company wrote. this blog. the post was published today. “This grant is a step to create democratic processes … to control superintelligence.”
Through grants from the non-profit organization OpenAI, OpenAI hopes to create a process that reflects the Platonic ideal of democracy: a “broadly representative” group of people exchanging opinions, engaging in “informed” discussions, and ultimately deciding the outcome. transparent process. decision making process.
Ideally, OpenAI says, the process will help answer questions like “Under what conditions should AI systems evaluate or criticize public figures, given the different opinions of different groups about these figures?” and “How should controversial views be represented in AI publications?”
“The main goal of this grant is to promote innovation in processes – we need better democratic methods to govern AI behavior,” OpenAI writes. “We believe that decisions about AI behavior should be based on diverse perspectives that reflect the public interest.”
– Our correspondent