The original plan to build FAI in secret and have it take over the world didn’t work out, and was unrealistic to begin with. But if economic competition is here to stay, that means AIs that are good toward humans by nature (not out of profit motive) can exist among others, and can be built using the same tools as for-profit AIs. In the long run maybe for-profit AIs will win out or maybe they won’t, but having some FAIs will at least give us a fighting chance.
To achieve that, we might not need very wide cooperation, but we do need some new organizational forms. OpenAI was started by believers and had good stated intentions, but they worked on winner-take-all logic, took lots of money, and fell from the side of good. What kind of organization would be best suited to work toward FAI as one-of-many, today?
I don’t know if there are official definitions, but to me, the connotation of “aligned AI” is something that’s aligned to the user. For example, an aligned AI working for TurboTax will happily help it lobby the government to keep tax filing difficult. That’s the kind of AI that corporations are trying to build and sell now. While the connotation of “friendly AI” to me is more about being aligned to the concept of good in general, without privileging the user. It probably needs to be built by a non-market organization.
Friendly AI.
The original plan to build FAI in secret and have it take over the world didn’t work out, and was unrealistic to begin with. But if economic competition is here to stay, that means AIs that are good toward humans by nature (not out of profit motive) can exist among others, and can be built using the same tools as for-profit AIs. In the long run maybe for-profit AIs will win out or maybe they won’t, but having some FAIs will at least give us a fighting chance.
To achieve that, we might not need very wide cooperation, but we do need some new organizational forms. OpenAI was started by believers and had good stated intentions, but they worked on winner-take-all logic, took lots of money, and fell from the side of good. What kind of organization would be best suited to work toward FAI as one-of-many, today?
At this point, how would you define or characterize Friendly AI? Do you consider “aligned AI” to be a completely separate thing?
I don’t know if there are official definitions, but to me, the connotation of “aligned AI” is something that’s aligned to the user. For example, an aligned AI working for TurboTax will happily help it lobby the government to keep tax filing difficult. That’s the kind of AI that corporations are trying to build and sell now. While the connotation of “friendly AI” to me is more about being aligned to the concept of good in general, without privileging the user. It probably needs to be built by a non-market organization.