Re:A.I. Companies Agree to Safeguards After Pressure from th...
by danShumway on Friday July 21, 2023 @12:07PM
(#36815538)
There are good (or at least sincere) actors talking about AI regulation. I don't agree with all of them, but I believe they're trying to help solve problems that they do believe actually exist.
OpenAI is not in that group. OpenAI is looking for regulatory capture and propaganda about their AI capabilities, that's it. It is not worth taking them at their word when they talk about AI safeguards. Same with companies like Meta/Facebook, same with Google.
It makes no sense to ask these companies to be in charge of safeguards. Not only are the safeguards they propose likely to be ineffective; in some cases (watermarks, limits on what is and isn't "too powerful") they're likely to be actively harmful, entrenching corporate power and making it even easier for those companies to abuse AI to harm ordinary people.
OpenAI is not in that group. OpenAI is looking for regulatory capture and propaganda about their AI capabilities, that's it. It is not worth taking them at their word when they talk about AI safeguards. Same with companies like Meta/Facebook, same with Google.
It makes no sense to ask these companies to be in charge of safeguards. Not only are the safeguards they propose likely to be ineffective; in some cases (watermarks, limits on what is and isn't "too powerful") they're likely to be actively harmful, entrenching corporate power and making it even easier for those companies to abuse AI to harm ordinary people.