Is there a broader term or cluster of concepts within which is situated the idea that human values are often downstream of decisions, not upstream, in that the person with the correct values will simply be selected based on what decisions they are expected to make (ie election of a CEO by shareholders). This seems like a crucial understanding in AI acceleration.
Sorry if I’m a little thick, but can you tell me if I’m on the right track: you’re looking for a term or concept which describes the phenomenon where shareholders may select a CEO not directly because of his moral values but because of his aptitude for decisions which increase shareholder value, and that the CEO’s decision making models would be in some way reflective of his moral values even though Moral Values are not the quality they’ve been selected for, right?
People normally model values as upstream of decisions. Causing decisions. In many cases values are downstream of decisions. I’m wondering who else has talked about this concept. One of the rare cases that the LLM was not helpful.
What do you mean “model values as upstream of decisions” what is an example of a value that could be modeled based on a type of common decision? Would “CEO fires 10% of workforce just before Christmas time” be a sensationalist example in that it appears, on the surface, to reveal the moral value “shareholder value is a greater than the peace of mind afforded to these families”?
Is there a broader term or cluster of concepts within which is situated the idea that human values are often downstream of decisions, not upstream, in that the person with the correct values will simply be selected based on what decisions they are expected to make (ie election of a CEO by shareholders). This seems like a crucial understanding in AI acceleration.
What do you mean by “values”? Do you mean moral values or do you mean what metrics they optimize for?
moral values
Sorry if I’m a little thick, but can you tell me if I’m on the right track: you’re looking for a term or concept which describes the phenomenon where shareholders may select a CEO not directly because of his moral values but because of his aptitude for decisions which increase shareholder value, and that the CEO’s decision making models would be in some way reflective of his moral values even though Moral Values are not the quality they’ve been selected for, right?
People normally model values as upstream of decisions. Causing decisions. In many cases values are downstream of decisions. I’m wondering who else has talked about this concept. One of the rare cases that the LLM was not helpful.
What do you mean “model values as upstream of decisions” what is an example of a value that could be modeled based on a type of common decision? Would “CEO fires 10% of workforce just before Christmas time” be a sensationalist example in that it appears, on the surface, to reveal the moral value “shareholder value is a greater than the peace of mind afforded to these families”?
“What causes your decisions, other than incidentals?”
“My values.”