What happened in this episode?
AJ Gentile interviews cognitive scientist Donald Hoffman about his long-running claim that human perception is not a veridical window onto reality but an adaptive interface shaped by natural selection. Hoffman argues—using evolutionary game theory language—that fitness payoffs generally do not preserve the underlying structure of the world, and therefore evolution does not favour ‘truth’ in perception. He extends this into a broader programme in which consciousness is treated as fundamental and modelled mathematically using Markov chains, with ‘agency’ expressed via recursive layers of policies and meta-policies and a proposed ‘trace logic’ linking perspectives.
Hoffman sharply criticises mainstream physicalist accounts of consciousness, saying neuroscience and computational approaches have not explained a single specific conscious experience, while acknowledging they have produced valuable descriptive findings about brains. He recounts early career work in aerospace software at Hughes Aircraft and describes joining Francis Crick’s private ‘Helmholtz Club’, which he says met monthly in Southern California to pursue consciousness research before it became a mainstream topic.
On current work, Hoffman claims he has shared tentative, unpublished proofs with collaborators aiming to derive special and general relativity—and eventually quantum rules such as the Born rule—from his framework; he repeatedly notes the work is not ‘real’ until peer-reviewed. The conversation also ranges into psychedelics and UAPs: Hoffman suggests DMT might alter the parameters of perception rather than merely produce hallucinations, and speculates that apparently ‘impossible’ UAP behaviour could reflect manipulation of the perceptual ‘headset’.
Late in the episode Hoffman gives an emotional account of a hospital crisis in which he texted his wife goodbye and says his heart rhythm abruptly normalised after he saw two unidentified figures at his bedside—an experience he says he has not previously discussed publicly.
