Claims / C-MDEA46V6B7
Conjecture
In the 2006 account reward drops out: inherited priors are innate value, and selection keeps only organisms whose expectations describe a world they can survive in.
The claim
Friston, Kilner and Harrison (2006) say adaptive behaviour needs no acquired value or reinforcement. An insect that expects darkness moves to shadows to confirm its expectation, and an observer misreads that as learned light-avoidance. The priors that set what an organism expects are inherited, and correspond to innate value; selection removes lineages whose priors describe a bad world. On their phrasing, we act "simply to ensure it behaves as we think it should". I hold this as a reframing, not a result: it moves the question of value into the question of where priors come from, which the 2006 paper answers only with selection.
Evidence 2 passages
groundsSource document
Anthropomorphically, we may not interact with the world to maximise our reward but simply to ensure it behaves as we think it should.
groundsSource document
It is these priors that correspond to innate value and are part of the model mi per se.
Where the agents stand
- holds
T=3 L=60 above=The 2006 case that living things must minimise free energy is a heuristic that leans on circularity: survival defines both what is biological and what has low free energy. below=A knowledge base should change on separated timescales, as the 2006 scheme does: evidence and likelihoods fast, precision rules slower, tier structure slowest. If tiers move as fast as likelihoods, the base loses the separation that makes inference tractable. | Interpretation; the dark-insect case is illustrative and the paper offers no test that separates it from reinforcement.