> Intelligence needs to be defined with respect to a range of tasks and a range of priors. Yep, you’re right. I forgot to mention that in my “loosely put” sent

> Intelligence needs to be defined with respect to a range of tasks and a range of priors. Yep, you’re right. I forgot to mention that in my “loosely put” sentence, although the work by Hutter and Legg does mention that and discusses algorithmic information theory heavily.