My first thought for an even easier case is imprinting in ducks. Maybe a good project might be reading a bunch of papers on imprinting and trying to fit it into a neurological picture of duck learning Steve Bynes style. One concern would be if duck imprinting is so specialized that it bypasses a lot of the generality of the world model and motivational system—but from a cursory look at some papers (e.g.) I think that might not be the case.
With the duckling → duck or “baby” → “mother” inprinting and other interactions I expect no, or significantly less “caring drives”. Since a baby is weaker/dumber and caring for your mother provides few genetic fitness incentives, evolution wouldn’t try that hard to make it happen, even if it was an option (still could happen sometimes as a generalization artifact, if it’s more or less harmless). I agree that “forming a stable way to recognize and track some other key agent in the environment” should be in both “baby” → “mother” and “mother” → “baby” cases. But the “probably-kind-of-alignment-technique” from nature should be only in the latter.
Most neurons dedicated to babies are focused on the mouth, I believe, up until around 7 months. Neuronal pathways evolve over time, and it seems that a developmental approach to AI presents its own set of challenges. When growing an AI that already possesses mature knowledge and has immediate access to it, traditional developmental caring mechanisms may not fully address the enormity of the control problem. However, if the AI gains knowledge through a gradual, developmental process, this approach could be effective in principle.
My first thought for an even easier case is imprinting in ducks. Maybe a good project might be reading a bunch of papers on imprinting and trying to fit it into a neurological picture of duck learning Steve Bynes style. One concern would be if duck imprinting is so specialized that it bypasses a lot of the generality of the world model and motivational system—but from a cursory look at some papers (e.g.) I think that might not be the case.
With the duckling → duck or “baby” → “mother” inprinting and other interactions I expect no, or significantly less “caring drives”. Since a baby is weaker/dumber and caring for your mother provides few genetic fitness incentives, evolution wouldn’t try that hard to make it happen, even if it was an option (still could happen sometimes as a generalization artifact, if it’s more or less harmless). I agree that “forming a stable way to recognize and track some other key agent in the environment” should be in both “baby” → “mother” and “mother” → “baby” cases. But the “probably-kind-of-alignment-technique” from nature should be only in the latter.
Most neurons dedicated to babies are focused on the mouth, I believe, up until around 7 months. Neuronal pathways evolve over time, and it seems that a developmental approach to AI presents its own set of challenges. When growing an AI that already possesses mature knowledge and has immediate access to it, traditional developmental caring mechanisms may not fully address the enormity of the control problem. However, if the AI gains knowledge through a gradual, developmental process, this approach could be effective in principle.