3 ms·

That's an interesting subject to think about and discuss. I think the existing legal system is already equipped with the precedents it needs to handle some aspe

That's an interesting subject to think about and discuss. I think the existing legal system is already equipped with the precedents it needs to handle some aspects of such situations. The idea of someone carrying out a crime via a proxy is not new; the main difference now is that the proxy may be an AI. I think something which is new though is the question of who should take responsibility for the AI's actions? If a crime is committed via a proxy, then both parties will be charged. So it doesn't make sense to entirely just waive away the AI's responsibility in the crime just as it doesn't make sense to acquit the person/proxy who carried out the crime on behalf of someone else. I think the answer is that the company(ies) which trained and/or operated the company behind the AI should also be held partially responsible for the crime. Otherwise we could end up with a situation whereby some companies are intentionally creating malicious bots which are trained to intentionally misinterpret benign-sounding language as orders of a more sinister nature in order to shield 'the customer' from criminal liability. I'm much more worried about the consequences of humans being under-punished rather than over-punished. Neglecting the role of AI companies seems like a sure way to bring about human extinction. We cannot allow all liability to be shifted to non-human entities. We need a better, more fine-grained way to attribute and punish shared and partial liability. We should think in a more fine-grained way. For example, if someone throws out a plastic Coca Cola bottle on the ground, who is responsible for littering? Obviously the person who threw away the bottle is mostly responsible but isn't Coca Cola company also slightly responsible for having made their bottles out of plastic (which doesn't break down easily; thus worsening the impact of the littering)? Surely they could just have sold metal cans and glass bottles.

xg15 3y ago
> The idea of someone carrying out a crime via a proxy is not new; the main difference now is that the proxy may be an AI. I thought that the concept discussed in the OP was slightly different (and possibly even more insidious) : The proxy is a human who carries out orders of (or covers for) an AI that is known to be unsafe. I.e. an autonomous car manufacturer may acknowledge that their AI sometimes makes errors which can lead to accidents - so they equip their cars with a steering wheel and mandate that a human driver must always be present who can takeover from the AI in dangerous situations. The problem is just that the human driver may not realistically be able to do that, because reaction times are too short, the AI releases control too late, etc. Yet, if the AI malfunctions and the driver doesn't manage to prevent the accident, the company can comfortably blame the driver and keep the AI out of the discussion.

xg15 3y ago

> The idea of someone carrying out a crime via a proxy is not new; the main difference now is that the proxy may be an AI. I thought that the concept discussed in the OP was slightly different (and possibly even more insidious) : The proxy is a human who carries out orders of (or covers for) an AI that is known to be unsafe. I.e. an autonomous car manufacturer may acknowledge that their AI sometimes makes errors which can lead to accidents - so they equip their cars with a steering wheel and mandate that a human driver must always be present who can takeover from the AI in dangerous situations. The problem is just that the human driver may not realistically be able to do that, because reaction times are too short, the AI releases control too late, etc. Yet, if the AI malfunctions and the driver doesn't manage to prevent the accident, the company can comfortably blame the driver and keep the AI out of the discussion.