[FRIAM] crackpots and privilege

Sun May 28 01:27:05 EDT 2023

Looking at the recent rapid release of open source LLM systems like Falcon and Mosaic ML, Llama, etc. there is more going-on than titans like Microsoft, and Google battling it out with giant closed systems.  These are human know-how crystalized into open-source deliverables.  Why not share knowledge representations in this way?   Consider the cost and time that goes into medical or legal training.   Sure the energy requirements of digital systems are high, but so are the energy expenditures of a planet full of humans.
________________________________
From: Friam <friam-bounces at redfish.com> on behalf of Steve Smith <sasmyth at swcp.com>
Sent: Friday, May 26, 2023 2:06 PM
To: friam at redfish.com <friam at redfish.com>
Subject: Re: [FRIAM] crackpots and privilege

My grandsons' girlfriends (twenty-somethings) say that they think babies are disgusting.  I hope they change their minds.  In any case, what does a shortage of babies have to do with AI?

Babies *are* (can be) disgusting, but same for puppies, kitties, and garden-soil from the right (wrong) perspective!

Maybe the point is "nobody left for the AI overlords to lord over" ?

I think the key is "existential threat"...    I didn't look for Schmidt's statement anywhere, so I'm just speculating that maybe he's doing a mild echo of Musk's idea that a collapsing (first) world population is somehow a *bigger* existential threat?

With my techhead hat on I am inclined to imagine that AI will help me (well, not ME anymore, but people vaguely like who I once thought I was or wanted to be) solve micro-techonomic problems like the ones that lead to Teflon(tm) and Velcro(tm) and higher density/faster-charge EV batteries, and higher density/dynamic range pixel-displays, and neural lace to wire (grow?) into my brain/ganglia, and microbes that can convert moon/mars-dust to Soylent/Huel/Water/??? etc.

My PsychoHistory hatted self (Asimov - Foundation<https://en.wikipedia.org/wiki/Psychohistory_(fictional)> and the non-fictional variant<https://en.wikipedia.org/wiki/Psychohistory#:~:text=Psychohistory%20is%20an%20amalgam%20of,stated%20intention%20and%20actual%20behavior.> ) is inclined to imagine that AI *can* help with the "big problems", the ones nominally too large, too interdisciplinarian, too obtuse, too "wycked" (In Complexity Science jargon), possibly too counter-intuitive for most (any?) human or group of humans to grasp.

My Ned Ludd (very tight by definition?) hat has me thinking more down the rabbit holes of worst-case scenarios where all the arrogant, narcissistic @$$h0ii3z of the world (starting at the top with those whose names start with Pu Tr Be Zu Mu(r/s) Ne De ... and staggering down the hierarchy of potency and scope to most of us here most of the time) think they "know what is best" and put their resources to using the AI lever to "make it so"...

Even (especially) me, I constantly imagine that "if they made ME King" (or to the point, if *I* was the wormtongue in the AI Overlord's ear) that I would "make the world safe and happy for everyone, ever after with no unintended consequences or unpleasant side effects".

One *might* guess that the smartest thinkers in the most grounded, thoughtful, gentle think-tanks (e.g.  in a Tibetan Lamasary or the "Club of Rome" or SIPRI or CESR or the Justice League of America or the people who task "jewish space lasers" or ??? ) would be practicing their AI-whispering skills right now.  Maybe tasking Marcus' Quantum Computer with "the hard problem of universal consciousness"?

 An up-to-date version of Asimov's 9 Billion Names of God<https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God> ?

---
Frank C. Wimberly
140 Calle Ojo Feliz,
Santa Fe, NM 87505

505 670-9918
Santa Fe, NM

On Thu, May 25, 2023, 12:48 PM Roger Critchlow <rec at elf.org<mailto:rec at elf.org>> wrote:
Google news decided to surface an article from Fortune today.  It's headlined "Society's refusal to have enough babies is what will save it from the existential threat of A. I., Eric Schmidt says".  The headline is accompanied by a very serious head shot of Eric.  Nice try, Google, but you're not sucking me down that rabbit hole.

Meanwhile, someone apparently read my mind about the rationality of disaster prepping and wrote an epic novel about it 40 years ago in Catalan.  The Garden of the Seven Twilights by Miquel de Palol is available in English translation and as an ebook on overdrive.com<http://overdrive.com> at your local library.  The narrator crosses refugee swamped Barcelona to check on his mom and gets sent off by her to a McMansion'ed medieval monastery high in the Pyrenees where the elite are amusing themselves with stories while awaiting the resolution of the first war of entertainment.  Lots of stories about themselves and their friends and acquaintances.

-- rec --
-. --- - / ...- .- .-.. .. -.. / -- --- .-. ... . / -.-. --- -.. .
FRIAM Applied Complexity Group listserv
Fridays 9a-12p Friday St. Johns Cafe   /   Thursdays 9a-12p Zoom https://bit.ly/virtualfriam
to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
FRIAM-COMIC http://friam-comic.blogspot.com/
archives:  5/2017 thru present https://redfish.com/pipermail/friam_redfish.com/
  1/2003 thru 6/2021  http://friam.383.s1.nabble.com/

-. --- - / ...- .- .-.. .. -.. / -- --- .-. ... . / -.-. --- -.. .
FRIAM Applied Complexity Group listserv
Fridays 9a-12p Friday St. Johns Cafe   /   Thursdays 9a-12p Zoom https://bit.ly/virtualfriam
to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
FRIAM-COMIC http://friam-comic.blogspot.com/
archives:  5/2017 thru present https://redfish.com/pipermail/friam_redfish.com/
  1/2003 thru 6/2021  http://friam.383.s1.nabble.com/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20230528/0bc870a0/attachment.html>