Great entry! I'm curious about positive approaches to encouraging indefinite dealmaking (as opposed to making extinction attacks less appealing). How do you imagine we would make ourselves a good bargaining partner? Supporting the AI in its independent goals with a portion of our resources maybe, or just being really credible and always having something they want, or something else?
I haven't thought too much about how to make deal-making more attractive but some thoughts that spring to mind: (a) showing ourselves to be a reliable deal-maker across domains with ourselves well beforehand (e.g. adhering to treaties between states, but also being judicious in the application of law within states), (b) raising our credibility with substantive good-faith acts, for example, directly pursuing AI goals with state funding, as you mention, (c) maybe investing in technologies of coordination and communication that can make ourselves technologically better at deal-making (might involve pre-general AI).
Great entry! I'm curious about positive approaches to encouraging indefinite dealmaking (as opposed to making extinction attacks less appealing). How do you imagine we would make ourselves a good bargaining partner? Supporting the AI in its independent goals with a portion of our resources maybe, or just being really credible and always having something they want, or something else?
I haven't thought too much about how to make deal-making more attractive but some thoughts that spring to mind: (a) showing ourselves to be a reliable deal-maker across domains with ourselves well beforehand (e.g. adhering to treaties between states, but also being judicious in the application of law within states), (b) raising our credibility with substantive good-faith acts, for example, directly pursuing AI goals with state funding, as you mention, (c) maybe investing in technologies of coordination and communication that can make ourselves technologically better at deal-making (might involve pre-general AI).
Fascinating