There seems to be a number of different possible approaches to building a Friendly AI, each with their own open problems. For example, we could design a system of uploads with safety checks and then punt all other questions to them, or figure out how we solve confusing philosophical problems and program a de novo AI with those methods, or solve them ourselves and just code up a decision process together with a set of preferences. Does this effort to define open problems assume a particular approach?
I will try to be clear as I go along which problems are relevant for which scenarios. Also, solving certain problems will mean that we don’t have to solve other problems. The trouble is that we don’t know which scenarios will play out and which problems we’ll solve first.
There seems to be a number of different possible approaches to building a Friendly AI, each with their own open problems. For example, we could design a system of uploads with safety checks and then punt all other questions to them, or figure out how we solve confusing philosophical problems and program a de novo AI with those methods, or solve them ourselves and just code up a decision process together with a set of preferences. Does this effort to define open problems assume a particular approach?
I will try to be clear as I go along which problems are relevant for which scenarios. Also, solving certain problems will mean that we don’t have to solve other problems. The trouble is that we don’t know which scenarios will play out and which problems we’ll solve first.