[FRIAM] /Topic Latent in: Latent Topics was: enough sleep?

[glen∈ℂ]

On 4/11/19 2:38 PM, Marcus Daniels wrote:
> I find this kind of evidence unsatisfactory.    How people act as individuals or in groups says nothing about how an AI might function as individuals or in groups.   It's merely an inventory of flaws and idiosyncrasies of our species.

Well, of course the implications of the results in that paper are not generalizable across species, much less to AI or across planets.  But we don't have any evidence that identity or entitlement exist anywhere other than humans, either. What's more likely is that they are like consciousness, a symptom of deeper causes, where those deeper causes *are* shared with other species.  ... like fish who recognize themselves in mirrors. If cross-species mind-reading has any validity, then it seems reasonable that animals might also experience some sort of "in the moment" flow where their self-referencing executive "egos" wane.  And if that flow could be shown to help schools, packs, and gaggles "bond" in some sense, then that paper is a tiny bit of added validation, meaningless on its own.