How to Build Trust in Artificial Intelligence
The OpenAI drama shows that companies can't police themselves. Government and nonprofits need help to ensure the technology benefits humanity.
Last month, OpenAI fired and then rehired CEO Sam Altman. When the dust settled, Open AI emerged looking even more like a for-profit corporation than a nonprofit, with a board that is indistinguishable from those of other tech startups—right down to the lack of diversity. This is a clear departure from the company’s 2015 origins as a nonprofit research group whose mission reflected a widely shared realization—that artificial intelligence is a powerful tool poised to impact us socially, economically and politically, and therefore needs guardrails.
But who should establish these guardrails? The answer is unclear. Companies will serve the public good when it aligns with profit. This is not a condemnation; it’s simply a fact. For those of us who worked at Twitter, the legal term “fiduciary responsibility” became a familiar phrase as we saw the digital town square sold to the highest bidder—a man we all knew would use the platform to amplify his personal and political perspectives.