OpenAI says it is launching a program to award ten $100,000 grants to fund experimentation in establishing a democratic process for deciding what rules AI systems should follow — “within the limits set by law.”
The launch of the grant program comes on the heels of OpenAI’s call for an international regulatory body for AI, similar to that for nuclear power. In their proposal for such a body, OpenAI co-founders Sam Altman, Greg Brockman and Ilya Sutskever argued that the pace of innovation in AI is so rapid that we cannot expect existing authorities to adequately rein in the technology – a sentiment announced today also catches the eye.
Specifically, OpenAI says it wants to fund individuals, teams and organizations to develop proof-of-concepts for a “democratic process” that could answer questions about guardrails for AI. The company wants to learn from these experiments, it says, and use them as the basis for a more global — and ambitious — process in the future.
“While these initial experiments (at least for now) are not intended to be binding on decisions, we hope they explore decision-relevant questions and build new democratic tools that can more directly inform future decisions,” the company wrote in a blog. message published today. “This grant is a step towards establishing democratic processes for the oversight of… superintelligence.”
With the grants, provided by OpenAI’s nonprofit organization, OpenAI hopes to establish a process that reflects the Platonic ideal of democracy: a “broadly representative” group of people who exchange opinions, participate in “intentional” discussions, and ultimately decide on an outcome through a transparent decision-making process. Ideally, OpenAI says, the process will help answer questions such as “Under what conditions should AI systems condemn or criticize public figures given groups’ differing opinions on those figures?” and “How should contested views be reflected in AI output?”
“The primary purpose of this grant is to promote innovation in processes – we need improved democratic methods to direct AI behavior,” OpenAI writes. “We believe that decisions about how AI behaves should be shaped by different perspectives that reflect the public interest.”
In the announcement post, OpenAI suggests that the grant program is completely separate from its commercial interests. But that’s a bit of a hard pill to swallow given Altman’s recent criticism of proposed AI regulation in the EU. The timing also seems striking, following Altman’s appearance last week before the US Senate Judiciary Committee, where he argued for a very specific flavor of AI regulation that would have minimal effect on OpenAI technology as it exists today.
But even if the program turns out to be selfish, it’s an interesting direction to take AI policymaking (albeit a copy of the EU’s effort in some obvious ways). For example, I’m curious to see what kind of ideas for “democratic processes” emerge – and which applicants OpenAI ultimately chooses.
People can apply for the OpenAI grant program starting today – the deadline is June 24 at 9pm PDT. Once the application period closes, OpenAI will select ten successful recipients. Recipients must present a draft involving at least 500 participants, publish a public report on their findings by October 20, and open source the code behind their work.