I agree completely. How do you ‘objectively’ define success? Its impossible to unmarry language from the human (subjective) concepts that they represent. In it’s purest sense, for AI to be objective, it would have to have never been created or the least bit influenced by human intelligence at all. But you wouldn’t want to encounter such an AI, I suspect.
Let’s also just note here my opinion, again, for the record, that what most people call Artificial Intelligence is really just adaptive programming. It’s not true AI that can not only adapt it’s limited processes, but can alter it’s own programming, scope, and even ‘grow’ or develop totally new functions
By the way, were the different examples of historical success that you mentioned the Roman Empire versus 20th century United States?
Actually when I posted the comment I was thinking in terms of the US vs. Haiti.
But they're all on the table.