[FRIAM] Philosophy, Ethics, Academia (PEA)

[⛧ glen]

Excellent point, assuming I'm grokking it. The Proof of Stake v PoW conversation touched on it. But my broken mind keeps hearkening back to Wolpert's limit that there can be only one most accurate model of the universe that's within the universe. I don't know if he's (or anyone else has) derived the comprehensions or power set of the lesser accurate models of the universe. 

It seems to me the question you pose is the extent to which such comprehension can be flattened/democratized.  Improvements like PoS over PoW help, but are still limited in their power to express the whole polity, way better than "price". But, banally, they're still only available to the Morlocks, opaque to the Eloi, in the same sense as the Blackrock funds pushing big pharma to develop vaccines without concern for profits.

Can we come up with a comprehension of ALL available sub-models of the world that maximizes the availability of that meta-model to as many stakeholders as possible? I'm optimistic we can, given a closed universe. But what if the universe is open?

What if Wolpert's result implies that, from minute to minute, the most accurate model changes because the universe changes ... and the possible comprehensions of sub-models also changes?

And yes, I see the 3 Constitutions (liberal, conservative, libertarian) as attempts to wiggle around our current one, all of which are sub-models of the world based on some form of social justice. But, as with our conversation about Originalism, to what extent will they remain accurate as the world changes?

To round out this post, my targeting of myopic cartoons like anarch-capitalism isn't an attempt to discourage work on such problems. It's an attempt to police ourselves, those of us who do think it can be done (at least in a piecewise stable universe). Anarcho-capitalists and many libertarians seem, to me, to be arguing in bad faith (in Sartre's sense). But anarcho-syndicalists don't seem that way to me. I'm surely biased, of course.

Thanks for bringing so many contexts together into a theme! I don't feel like I've addressed the 1-way algorithm question directly. But if I could answer some questions about comprehensions of sub-models, then I could talk more coherently about how to fix one comprehension over another.

I'm also worried that everything I think at 2am is nonsense anyway. Sorry if that's true.

On April 1, 2021 1:38:16 PM PDT, David Eric Smith <desmith at santafe.edu> wrote:
>Glen,
>
>I like the fact that you word the following so as to suggest it’s
>impossible by tautology:
>
>> How can there *ever* be any form of capitalism that's NOT "crony"? If
>there's a state, the state will be ... [cough] ... leveraged to gain
>asymmetric power.
>
>I like it because the implication that something looks impossible makes
>us take seriously that it is at least hard.
>
>
>Can I propose that this is a good starting point for defining some
>aspects of democracy in aspirational terms, as a performance criterion
>for which one then seeks solutions.
>
>There are many things that look intuitively like they should be
>impossible by tautology, and which turn out to have useful solutions
>that took work to find or build.
>
>1. How can there be securities markets with large scope for autonomous
>action but that are strongly effective in blunting the exercise of
>power by wealthy actors to manipulate prices, extract information about
>counterparties, or otherwise go outside the domain of contract?  But
>the whole Fama-French efficiency assertion (which is wrong, but for
>more subtle reasons) gives a good quantification of how many such
>actions the design of securities markets does blunt.
>
>2. How can there be a cognitive discipline built around the moral of
>the Emperors New Clothes (the aspiration of science).  Anyone should be
>able to overturn a position no matter how firmly held, by providing
>evidence of its error, and ballot-stuffing should not be strong enough
>to overcome that lever.  To the extent that scientific claims are taken
>to have a reliability that is different from just whoever’s opinion
>about whatever, they measure whether this has been achieved.
>
>3. How can there be an encryption algorithm that provides easy
>encryption in public but effectively unaffordable decryption by the
>same agents?  How can there be proofs of identity that are cheaply
>verifiable but effectively unspoofable?  Etc.  The world that was
>opened by 1-way algorithms.
>
>
>I am comfortable taking as a starting framing that the problem with the
>traditional formalization of economics is as the study of problems of
>allocation in the arena where power is excluded from operating. 
>Mathematically one can put some constructions behind that, but then to
>what areas of life do we think the absence of power is at all an
>adequate approximation?  (Your question above, but I think there are
>some for which the cartoon is quite helpful; auction design is a nice
>example of a problem with many good solutions, arrived at technically.)
>
>That leaves to Political Science (or whatever mash of disciplines) the
>study of power, its uses, consequences, nature, or whatever.  
>
>The two are then bound, in a formalism Shubik used to use enough times
>that it is tattooed in my traces — the Economy exists within the Polity
>and the Society.
>
>So we can ask:  to what extent can rules of governance be designed to
>have a 1-way character in blunting certain forms of the exercise of
>power, enough to make the polity a safe holder of power that can
>constrain its use by other actors?  Are there 1-way aspirations that
>are provably unreachable?  By trying to prove nonexistence theorems
>that aren’t true, do we gain insights that might lead to the discovery
>of 1-way algorithms we aren’t currently using?  People seem to think
>the Magna Carta and the US Constitution were significant contributions
>(to several aims, but I think that was one).  I take your posts on the
>communities looking at constitutional reform and redesign as efforts to
>find more.
>
>This all seems a good framing to me.
>
>Eric

-- 
glen ⛧