[FRIAM] Free Will in the Atlantic

[uǝlƃ ↙↙↙]

I don't understand why it matters that the reducing function is also a function. Why keep harping on that? It's machines everywhere. Big deal. But as Steve and Gary point out re: dancing robots, the reducing function is also complicated. When someone who's used to seeing CGI-animations *or* people trying to dance awkwardly because there's social pressure to do so but don't *feel* whatever they're dancing to, the reflective layer is, following Ashby, at least as complicated as the machine doing the thing.

So, the reflective layer truly is a covariate and can't be approximated out. The task is to build a machine that acts sufficiently like the extant machines (people) in exhibiting what we're calling free will.

On 4/5/21 9:48 AM, Marcus Daniels wrote:
> That agency could be reflective and be correlated with different social outcomes is just another curious covariance matrix.   It’s the attribution of causation from that reflective layer that is pulled out of thin air, because that reflective layer is just another machine.

-- 
↙↙↙ uǝlƃ