[FRIAM] Getting You Libertarians' Goats

[uǝlƃ ↙↙↙]

Well, the point I was making is that the physics Marcus is claiming must be exercised/computed *without* shared structure will *usually* result in shared structure. I.e. it'll be a rare thing for any component to be completely alien to another component. 

Now, "recognition" of any extant common structure is a 2nd order structure. So any 1st order variation/error between the 2 (multi-order) structures may well compound and produce larger variation/error in that recognition of shared structure.

And it's that combinatorial *expansion* of error I'm relying on to assert the rarity of disjoint components (that don't share any structure). That we can simulate/model each other AT ALL indicates shared structure.

And this is how I'm explaining the "unreasonable effectiveness" of these higher order behaviors (like _mind_).

Whether it's crisply bidirectional (isomorphic, recursively invertible) or not is a question we could talk about. But, yes, it has to be at least a little bit bidirectional. That's what "shared structure" means. A can't be similar to B without B also being similar to A, even if the 1st "similar" is slightly different from the 2nd "similar".

My only suggestion for downward causation lies in the higher order patterns canalizing the space, constraining the variation of the lower order processes. It's a different kind of "causation" than upward causation. Upward is driven, generative. Downward is constraint, restrictive ... like a flow-formed arroyo that is both caused by and causes flow patterns.

On 10/28/20 12:14 PM, thompnickson2 at gmail.com wrote:
> I would go with this so long as we agree it's bidirectional.  I learn who I am from watching you and you learn who you are from watching me and we both learn who we are by watching them.  
> 
> 
> -----Original Message-----
> From: Friam <friam-bounces at redfish.com> On Behalf Of u?l? ???
> Sent: Tuesday, September 15, 2020 7:17 AM
> To: friam at redfish.com
> Subject: Re: [FRIAM] Getting You Libertarians' Goats
> 
> I'd argue that, as long as there's common structure between the agents building and using the control layer models, then there will be shared concepts in and of those models. And it's not *only* the Libertarian who claims there's shared structure between the builders/users. As I've tried to *ask* w.r.t. to inter-species mind-reading, don't most complex animals "assume" some shared structure, not only between them and their kin, but between them and their predators/prey and symbiotic species? And if your dog does it, then it's likely Capitalists and Socialists also do it.
> 
> Of course, the extent to, and rate at, which such shared structures change through time is the gist of this conversation.
> 
> On 9/14/20 8:16 PM, Marcus Daniels wrote:
>> Using statistical mechanics to inspire a stable and universal functional form that evolves in time is one way to make a model of social systems.   But even with that for model of the physical world, there are many possible models for control systems that could layer on top of it.   If there are no shared concept types in these different models, there's nothing to do but go back to simulating the physics to determine what could happen next.   Simulating these physics takes energy that is of no discernable value to users of any one model so at some point there will be conflict over that energy.    The Libertarian claims that there is something in common between the users of these models, but it is nothing more than story that serves her purposes.   There is no reason not to violate her sovereignty if the reward/risk is acceptable.  

-- 
↙↙↙ uǝlƃ