[FRIAM] Murdoch and Trump

[uǝlƃ ☣]

The problem with Marcus' question is its 2 types of closure. 1) communication, reason, and action are separable. But the question convolves them. And 2) any instance of communication, reason, or action won't be complete. (I'm reminded of Wolpert's paper: https://arxiv.org/abs/0708.1362)

And this (also) hearkens back to the scientific use of incomprehensible models. The sentiment from Siegenfeld that the objects/mechanisms in a model can be obtuse without preventing us from learning about their behavior, combined with Pieter's mention of Deutsch, got me thinking again about these preemptive closures. Someone like Deutsch, whose cognitive abilities seem to be more expansive than most, will tend to over-estimate the cognitive abilities of others (cf Dunning-Kruger), lending to a tendency to be delusionally optimistic about human progress ... or at least a tendency in hindsight to believe we've done all this intentionally ... instead of stumbling like idiots into a lucky sweet spot. 

The work on "zero-intelligence agents" can be used to make the argument that these brainy people, with 6σ expansive intellects, actually *hinder* our progress. The progress we've made consists largely of us morons taking stunted action, with stunted communication, and stunted reason. I've made this argument before when criticizing "Effective Altruism" and some other trends in the "rationalist" community.

Seen another way, a better example of an obtuse model than Deutsch is Feynman, because he was such a great teacher. I think it's fair to say that *we* don't understand the internal mechanisms that composed the animal we call "Richard Feynman". Yet, he helped us make scientific progress. I expect some might claim that Feynman wasn't a *model* of, say, quantum structures. But that claim forgets what computer programmers know in a deep sense, that has-a relations and is-a relations can produce the same result. Whether Feynman has-a model of the quantum or is-a model of the quantum is irrelevant to the large-scale, collective wave of progress.

And so, in addition to my 2 ways of making scientific progress with obtuse models (parallax and expressibility), we have this 3rd way implied by Marcus and bolstered by Pieter. It would be interesting if there were a scale-free distribution of intellect expansiveness (with a small number of big ones like Deutsch down to lots of morons like ants or worker bees). My guess is the complexity fanbois should be arguing that it's that network *structure* that leads to the progress, rather than focusing on any 1 scale.

On 1/21/20 3:05 PM, Marcus Daniels wrote:
> How about one step back:  Are we involved in a collective process where communication, reason, and action are possible? 
-- 
☣ uǝlƃ