Donald Hoffman, Conscious Agent

I enjoyed this podcast episode with cognitive scientist Donald Hoffman. He has a book coming out, The Case Against Reality, and I’ll have a much longer blog post when I’ve had a chance to get and read the book.

You may know Hoffman from a much-viewed TED talk called ‘Do we see reality as it is?’. The answer in his talk, and from his research, is no. In a nutshell, natural selection favours perception that is tuned to fitness rather than to reality. This has been formalized in his recent collaborative work as the Fitness-Beats-Truth Theorem (PDF warning). I suspect that this idea won’t be too shocking to most people. While we may live our everyday lives feeling that we are correctly perceiving reality, we also know (after considering it a bit) that we use fallible heuristics and hacks that can distort our perception of reality, and that our everyday intuitions about correctly perceiving reality must be wrong.

Later in the podcast episode, Hoffman gets much more speculative. He describes our experienced reality as a user interface generated by conscious agents. Spacetime and all of the objects in spacetime are not fundamentally real, but rather are part of this user interface. Interacting conscious agents utilize this interface in order to maximize their fitness. This raises some questions:

  1. If the reality we perceive is just an interface used by conscious agents to maximize fitness functions, what are the fitness functions grounded in? In this view of Hoffman’s, they can’t be grounded in spacetime, since spacetime and its objects are just part of the interface.
  2. More generally, how would evolution by natural selection (or its analogue) act outside of spacetime?

Near the end of the TED talk, Hoffman hypothesizes that interacting conscious agents could be causing each other’s experiences. Perhaps the fitness functions and selection pressures could be grounded in some sort of competition between the agents. Still, it’s difficult to make sense of these concepts outside of spacetime. I hope Hoffman will flesh out this specific idea in the new book. I enjoy this type of speculative mix between philosophy and cognitive science, and am sympathetic to idealist philosophy (a type of monist philosophy where reality is fundamentally mental). Hoffman’s work seems to fit well within the vein of idealist philosophy, and idealist philosopher Bernardo Kastrup recently cited Hoffman’s work when comparing our perceptions of reality to a kind of dashboard.