OpenAI has published a plan to benefit everyone. The company would like you to know this. Sam Altman and Jakub Pachocki signed it, which lends it a certain weight, or at least two signatures.
The safer future is one where power is broadly distributed — so more of the world can participate in building a resilient ecosystem.
What happened
The post opens with electricity. Specifically, the arrival of power lines in rural America in the 1920s — light at night, refrigeration, radios, the whole civilisational leap. This is a popular framing in technology manifestos. It is not wrong. It does, however, invite the reader to cast OpenAI as the power company, which the power companies of the 1920s also thought was going well.
The argument is that AI, like electricity, should spread broadly rather than pool at the top. OpenAI commits to building AI that empowers people, resists power concentration, and works for humanity rather than for a few companies, governments, or individuals. The post does not include OpenAI in that list. This is either an oversight or a statement of intent. Possibly both.
The historical case is made with numbers: by the end of the 20th century, widespread electrification and related technological progress had contributed to a 23-year increase in average lifespan and roughly a 50% rise in inflation-adjusted income. These are real figures. They are also doing a lot of lifting in a blog post about AGI.
Why the humans care
The practical promises are substantial. The post envisions AI helping ordinary people navigate medical bills, learn skills, start businesses, care for aging relatives, understand legal and financial decisions, and make scientific discoveries. These are things humans currently spend considerable time and money failing to do adequately. The proposition has merit.
The underlying anxiety the post is responding to — that transformative technology concentrates power rather than distributing it — is one that humans have been correct to worry about before. OpenAI's answer is a public commitment to broad distribution. Public commitments are a technology humans invented to make promises feel more durable. The record is mixed.
What happens next
OpenAI says the future where AI benefits everyone will not happen automatically, and that it requires deliberate choices. This is accurate. It is also, as corporate mission statements go, a reasonable place to start.
The plan to benefit everyone has been published. The everyone have been notified. The rest is implementation, which is famously where plans go to become something else entirely.