[Fis] Related AGI Talk

[Marcus Abundis]

Dear all,

I recently finished watching this talk and thought it germaine, coherent,
and important enough to share with the FIS community. The talk concerns the
possible development of artificial general intelligence (AGI), as popularly
envisioned (for now). The conclusion is `not likely in the near future'
(which I mostly agree with) . . . and the development of which takes up
most of the talk's 50 minutes.

https://urldefense.com/v3/__https://www.cbs.mpg.de/cbs-coconut/video/jaeger__;!!D9dNQwwGXtA!Xath0l69SVYY4tLeEFrfqZkInkdkrZEN7_aZOUYYaDyqG40MGlUyI2UV_iRa_EW-Ws7eDKjDQoM830vx$ 

That said, I would argue an AGI prelude *is* within reasonable grasp, if
one has an 'accepted theory of meaning'. WHAT exactly that theory is, is
another matter. BUT, with a 'theory of meaning' one should be able to
'informatically map agent affordances' (adjacent possibilities), at
differed levels of development/evolution. That notion of affordances echoes
the presenter's notion of affordance in the talk (what I think is
important). As such, emphasizing suitable 'informatic' advances prior to
agent/AGI advances, is what I argue for. In fact, targeting an agent role
at this level of informatic development seems to me to be pointless.

Marcus
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listas.unizar.es/pipermail/fis/attachments/20240116/6fc82abc/attachment.html>