This post seems mostly reasonable in retrospect, except that it doesn’t specifically note that it seems unlikely that voluntary RSP commitments would result in AI companies unilaterally pausing until they were able to achieve broadly reasonable levels of safety. I wish the post more strongly emphasized that regulation was a key part of the picture—my view is that “voluntary RSPs are pauses done right” is wrong, but “RSPs via (international) regulation are pauses done right” seems like it could be roughly right. That said, I do think that purely voluntary RSPs are pretty reasonable and useful, at least if the relevant company is transparent about when they would proceed despite being unable to achieve a reasonable level of safety.
As of now at the start of 2025, I think we know more information that makes this plan looks worse.[1] I don’t see a likely path to ensuring 80% of companies have a reasonble RSP in short timelines. (For instance, not even Anthropic has expanded their RSP to include ASL-4 requirements about 1.5 years after the RSP came out.) And, beyond this, I think the current regulatory climate is such that we might not get RSPs enforced in durable regulation[2] applying to at least US companies in short timelines even if 80% of companies had good RSPs.
The EU AI act is the closest thing at the moment, but it might not be very durable as the EU doesn’t have that much leverage over tech companies. Also, it wouldn’t be very surprising if components of this end up being very unreasonable such that companies are basically forced to ignore parts of it or exit the EU market.
I wish the post more strongly emphasized that regulation was a key part of the picture
I feel like it does emphasize that, about as strongly as is possible? The second step in my story of how RSPs make things go well is that the government has to step in and use them as a basis for regulation.
I think the post could directly say “voluntary RSPs seem unlikely to suffice (and wouldn’t be pauses done right), but …”.
I agree it does emphasize the importance of regulation pretty strongly.
Part of my perspective is that the title implies a conclusion which isn’t quite right and so it would have been good (at least with the benefit of hindsight) to clarify this explicitly. At least to the extent you agree with me.
This post seems mostly reasonable in retrospect, except that it doesn’t specifically note that it seems unlikely that voluntary RSP commitments would result in AI companies unilaterally pausing until they were able to achieve broadly reasonable levels of safety. I wish the post more strongly emphasized that regulation was a key part of the picture—my view is that “voluntary RSPs are pauses done right” is wrong, but “RSPs via (international) regulation are pauses done right” seems like it could be roughly right. That said, I do think that purely voluntary RSPs are pretty reasonable and useful, at least if the relevant company is transparent about when they would proceed despite being unable to achieve a reasonable level of safety.
As of now at the start of 2025, I think we know more information that makes this plan looks worse.[1] I don’t see a likely path to ensuring 80% of companies have a reasonble RSP in short timelines. (For instance, not even Anthropic has expanded their RSP to include ASL-4 requirements about 1.5 years after the RSP came out.) And, beyond this, I think the current regulatory climate is such that we might not get RSPs enforced in durable regulation[2] applying to at least US companies in short timelines even if 80% of companies had good RSPs.
I edited to add the first sentence of this paragraph for clarity.
The EU AI act is the closest thing at the moment, but it might not be very durable as the EU doesn’t have that much leverage over tech companies. Also, it wouldn’t be very surprising if components of this end up being very unreasonable such that companies are basically forced to ignore parts of it or exit the EU market.
I feel like it does emphasize that, about as strongly as is possible? The second step in my story of how RSPs make things go well is that the government has to step in and use them as a basis for regulation.
I think the post could directly say “voluntary RSPs seem unlikely to suffice (and wouldn’t be pauses done right), but …”.
I agree it does emphasize the importance of regulation pretty strongly.
Part of my perspective is that the title implies a conclusion which isn’t quite right and so it would have been good (at least with the benefit of hindsight) to clarify this explicitly. At least to the extent you agree with me.