Certainly you still need legal accountabilityāwhy wouldnāt we have that? If we solve alignment, then we can just have the AIās owner be accountable for any law-breaking actions the AI takes.
I agree that that is a very good and desirable step to take. However, as I said, it also incentives the AI-agent to obfuscate its actions and intentions to save its principal. In the human context, human agents do this but are independently disincentivized from breaking the law they face legal liability (a disincentive) for their actions. I want (and I suspect you also want) AI systems to have such incentivization.
If I understand correctly, you identify two ways to do this in the teenager analogy:
Rewiring
Explaining laws and their consequences and letting the agentās existing incentives do the rest.
I could be wrong about this, but ultimately, for AI systems, it seems like both are actually similarly difficult. As youāve said, for 2. to be most effective, you probably need āAI police.ā Those police will need a way of interpreting the legality of an AI agentās {āmentalā state; actions} and mapping them only existing laws.
But if you need to do that for effective enforcement, I donāt see why (from a societal perspective) we shouldnāt just do that on the actorās side and not the āpoliceāsā side. Baking the enforcement into the agents has the benefits of:
Not incentivizing an arms race
Giving the enforcerās a clearer picture of the AIās āmental stateā
I want (and I suspect you also want) AI systems to have such incentivization.
Not obviously. My point is just that if the AI is aligned with an human principal, and that human principal can be held accountable for the AIās actions, then that automatically disincentivizes AI systems from breaking the law.
(Iām not particularly opposed to AI systems being disincentivized directly, e.g. by making it possible to hold AI systems accountable for their actions. It just doesnāt seem necessary in the world where weāve solved alignment.)
I donāt see why (from a societal perspective) we shouldnāt just do that on the actorās side and not the āpoliceāsā side.
I agree that doing it on the actorās side is better if you can ensure it for all actors, but you have to also prevent the human principal from getting a different actor that isnāt bound by law.
E.g. if you have a chauffeur who refuses to exceed the speed limit (in a country where the speed limit thatās actually enforced is 10mph higher), you fire that chauffeur and find a different one.
(Also, Iām assuming youāre teaching the agent to follow the law via something like case 2 above, where you have it read the law and understand it using its existing abilities, and then train it somehow to not break the law. If you were instead thinking something like case 1, Iād make the second argument that it isnāt likely to work.)
I agree that that is a very good and desirable step to take. However, as I said, it also incentives the AI-agent to obfuscate its actions and intentions to save its principal. In the human context, human agents do this but are independently disincentivized from breaking the law they face legal liability (a disincentive) for their actions. I want (and I suspect you also want) AI systems to have such incentivization.
If I understand correctly, you identify two ways to do this in the teenager analogy:
Rewiring
Explaining laws and their consequences and letting the agentās existing incentives do the rest.
I could be wrong about this, but ultimately, for AI systems, it seems like both are actually similarly difficult. As youāve said, for 2. to be most effective, you probably need āAI police.ā Those police will need a way of interpreting the legality of an AI agentās {āmentalā state; actions} and mapping them only existing laws.
But if you need to do that for effective enforcement, I donāt see why (from a societal perspective) we shouldnāt just do that on the actorās side and not the āpoliceāsā side. Baking the enforcement into the agents has the benefits of:
Not incentivizing an arms race
Giving the enforcerās a clearer picture of the AIās āmental stateā
Not obviously. My point is just that if the AI is aligned with an human principal, and that human principal can be held accountable for the AIās actions, then that automatically disincentivizes AI systems from breaking the law.
(Iām not particularly opposed to AI systems being disincentivized directly, e.g. by making it possible to hold AI systems accountable for their actions. It just doesnāt seem necessary in the world where weāve solved alignment.)
I agree that doing it on the actorās side is better if you can ensure it for all actors, but you have to also prevent the human principal from getting a different actor that isnāt bound by law.
E.g. if you have a chauffeur who refuses to exceed the speed limit (in a country where the speed limit thatās actually enforced is 10mph higher), you fire that chauffeur and find a different one.
(Also, Iām assuming youāre teaching the agent to follow the law via something like case 2 above, where you have it read the law and understand it using its existing abilities, and then train it somehow to not break the law. If you were instead thinking something like case 1, Iād make the second argument that it isnāt likely to work.)