Skip to main content
20 events
when toggle format what by license comment
Dec 6 at 13:03 comment added Olivier5 @JoWehler I haven't read chapter 7 of course, but I like the tittle. :-) how could the future possibly get written, and by whom?
Dec 6 at 8:39 comment added Jo Wehler @Olivier5 The quoted section from chapter 12 brings together some concepts from the philosophy of nature with the scientific explanations developed before. The leading idea is to introduce evolution as a phenomenon in time: It builds bottom-up organism with high-level structures which a posteriori provide a top-down causality. – Currently I’m not convinced by Mitchell’s “Chapter 7: The future is not written”. And still, I have to read some remaining sections of his book.
Dec 6 at 7:40 comment added Olivier5 @JoWehler How well did he manage in that chapter? Were you convinced by his approach, or did you find something missing there, in that last stretch?
Dec 5 at 22:04 comment added Annika @Olivier5 ^__^ lol
Dec 5 at 21:29 comment added Olivier5 @Annika It is a metaphor, like "upward causation". There's no elevator actually involved.
Dec 5 at 21:10 comment added Annika @Olivier5 yeah, I've always thought of downward causation as a sophisticated form of compatiblism
Dec 5 at 20:46 comment added Jo Wehler @Olivier5 Mitchell discusses top-down causation i.a. in chapter 12 "Free Will", section "Some philosophical issues".
Dec 5 at 20:41 history edited Jo Wehler CC BY-SA 4.0
added 239 characters in body
Dec 5 at 20:18 comment added Olivier5 That seems like a good book. Re. "downward causation", I thinks it's best understood as the causal power of systems and structures eg to achieve a certain level of resilience and self-sustainment.
Dec 5 at 20:05 comment added Ben Hocking I find the use of the word "agents" to be very evocative considering the current LLM landscape. There are LLM agents that take actions that aren't completely predictable (and are sometimes downright confounding) and are even governed by a parameter called "noise" (to increase or decrease randomness). Do these LLM agents have "free will" per Mitchell's arguments? If not, what would it take for them to have "free will"? (NB: I do not believe LLMs are "conscious" or have "free will", but they are unpredictable. The question is, are we conscious and have free will?)
Dec 5 at 19:54 comment added Annika +1 and accepted for the fascinating link to Mitchell — look forward to diving in
Dec 5 at 19:53 vote accept Annika
Dec 5 at 18:22 comment added Jo Wehler @Annika Restricted to some lines of comment I can not do better than recommending the about 12 chapters from Mitchell’s book. I for myself did profit enormously from this book. I hope not to fall back behind Mitchell’s insights in further discussions about understanding free will within the scientific frame of determinism. (2/2)
Dec 5 at 18:22 comment added Jo Wehler @Annika Mitchell’s book opened-up my eyes how barren it is to discuss free will solely with terms from philosophy (of nature) like determinism, random events, “I could have done other”, agency, will etc. Mitchell's new aspect includes how species developed during their ontogenesis by continuous adaption to their environment: Which systems and algorithms did they develop to control how their drives trigger their actions? And how does their organism evaluate the corresponding feedback and decide about its further action? (1/2)
Dec 5 at 14:47 comment added Annika but maybe I'm missing something -- I unfortunately don't have time to delve deeply into one scientist's theory right now but I'll take your word on the contents you've summarized. I also get suspicious when non-physicists (especially biologists) start talking about "information" and things that are many levels below their main area of study. To date, I don't think we even know of a way to test that we are seeing downward causation or the action of embodied information.
Dec 5 at 14:46 comment added Annika "....And the meaning of that information is embodied in the structure of the system itself, based on its history." --> that is perfectly fine, but it suggests, at best, downward causation from system to parts. This type of causation can also be perfectly deterministic and therefore obviate the need to any extra cause like "will"
Dec 5 at 14:45 comment added Annika I'm having a hard time seeing Mitchell's position as anything more than dressed up determinism. Appealing to "information" and "the system" still appear to be serving the roles as objective causes of behavior vs some internal will.
Dec 5 at 7:02 comment added Jo Wehler @NotThatGuy For more details please see his talk. As far as I'm concerned subsequently I read his book.
Dec 5 at 6:33 comment added NotThatGuy "compatibilism ... cannot explain how free will or indeed any form of real agency can possibly exist in a deterministic universe" - this objection is trivially resolved if one views compatibilism as merely a semantic distinction with strict determinism. This would leave Mitchell's objection itself as merely a semantic disagreement - he merely rejects the compatibilist definition of free will, and says it doesn't meet the definition of "real agency" (whatever that means).
Dec 5 at 6:20 history answered Jo Wehler CC BY-SA 4.0