I think I disagree (that generality is irrelevant (this is only the case, because nfl-theorems use unreasonable priors)). If your problem has “any” structure: your environment is not maximally random, then you can use Occam’s razor and make sense of your environment. No need for the “real world”. The paper on universal intelligence is great, by the way, if formalizing intelligence seems interesting.
To spell out how this applies to your comment: If you use a reasonable prior, then intelligence is well-defined: some things are smarter than others. For example, GPT3 is smarter than GPT2.
I think I disagree (that generality is irrelevant (this is only the case, because nfl-theorems use unreasonable priors)). If your problem has “any” structure: your environment is not maximally random, then you can use Occam’s razor and make sense of your environment. No need for the “real world”. The paper on universal intelligence is great, by the way, if formalizing intelligence seems interesting.
To spell out how this applies to your comment: If you use a reasonable prior, then intelligence is well-defined: some things are smarter than others. For example, GPT3 is smarter than GPT2.