Free will is a terrible name for it, and arguably inherently contradictory, like "involuntary fasting": it's a synonym for an uncaused cause.
What's lost in the process of burning that strawman, is that we can't ever have perfect knowledge of our environment or our nervous systems; and in a seeming paradox of determinism, a being who believes they have agency will behave differently than one who does not. Even if one accepts the model that we merely rationalize our subconscious choices (an imperfect model with some non-zero truth), the data from the outcome of that choice still feed into one's neural network and influence future choices, making the "free will software stack" a potentially valuable psycho-technology.
As a practical matter, I'm intrigued by the model that what we experience as consciousness, and the feeling of agency, really only has one knob to turn: what tiny fraction of our massive neural information network we orient the spotlight of our attention towards. That singular "decision" point, navigating what Buddhist cognitive scientist John Vervaeke calls the "salience landscape", has incredible potency on its own, iterated moment after moment for the entirety of a human lifespan.
"the data from the outcome of that choice still feed into one's neural network and influence future choices"
It seems to me that any time you have a process that feeds the output into the input, then you can easily have a divergence of trajectory whose precision exceeds anything that could be explicitly stored in a finite universe.
So it seems to me that free will must exist in the sense that a simple system, let alone a brain, can produce information from somewhere that can't come from the limited universe we seem to inhabit. Like, take the Mandelbrot set as a simple proof of concept. It may not have all the things in it that are in the universe, but it seems to have more resolution, more components, than can exist in reality. So if you used it to feed a process, it can add something to the universe as it affects reality.
What's lost in the process of burning that strawman, is that we can't ever have perfect knowledge of our environment or our nervous systems; and in a seeming paradox of determinism, a being who believes they have agency will behave differently than one who does not. Even if one accepts the model that we merely rationalize our subconscious choices (an imperfect model with some non-zero truth), the data from the outcome of that choice still feed into one's neural network and influence future choices, making the "free will software stack" a potentially valuable psycho-technology.
As a practical matter, I'm intrigued by the model that what we experience as consciousness, and the feeling of agency, really only has one knob to turn: what tiny fraction of our massive neural information network we orient the spotlight of our attention towards. That singular "decision" point, navigating what Buddhist cognitive scientist John Vervaeke calls the "salience landscape", has incredible potency on its own, iterated moment after moment for the entirety of a human lifespan.