(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
. . . . . . . . . .



AI as a Bogus Business... [1]

['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.']

Date: 2025-01-13

This started off as a comment to Imitative AI Is Not a Viable Business or Capitalism Is Not Rational but grew into something by itself… See above for context, something missing in so many AI models.

Imitative AIs have no model of the world. Therefore, they have no way of knowing what is reasonable and what is not, what is life-like and what is not.

You hit the nail on the head. AI has no “life” where such models can even be grown and matured and no amount of data will make that so. When I started working in the “Valley”, I was a staff programmer on a Stanford AI research project (SUMEX-AIM). Full disclosure, I was not an AI researcher/coder; I and other “system programmers” kept the timesharing/ARPAnet software running for them. But that was an “expert system”. They had a model, the experience and expertise painfully and slowly developed from interacting with real medical experts. The code just did what the experts repetitively did every day. That AI model had no clue about anything else such as that golf even existed let alone anything about how Dr. ____ played the game. But it was really good at repetitively regurgitating what a real life expert learned over a lifetime about one narrow but deep subject. And the constant refinement/update cycle was laboriously expensive, hence the pipe dream (cheap fix) of “large data” — get enough bullshit and you might make good manure.

Recall the “Imitation Game”, a film about Alan Turing. He called it so because it was a “test” as to whether a computer (didn’t exist really at the time but...) could imitate a person for a set of questions. He was brilliantly on to something but, as we see now, and I figured out the last time AI was big, imitation is a far cry from real. If we are to do the “Turing Test” more accurately now, it is not a question but an awareness, namely, when you reach for the reset button or the power plug, the machine screams, “Get your F%$king hand off my power switch/cable!!!!!”. That is the real test and that only comes from self awareness/consciousness which itself comes from a real, lived life, a biological/human one aware of its own death. Assuming that, with enough data, it will become so is, as called in the theology trade, a “faith statement”1. AI is reductionist thinking at its core, i.e., get enough parts and you have more than a whole but it is more accurately reductio ad absurdum. — which also describes the other part of your piece, the Sand Hill Road phenomenon2.

The article is correct about the other, capitalism, side too. Before transistors there were only vacuum tubes (and analog). That change was a big deal. It made Turing’s digital logic (a computer) possible. But putting a bunch of transistors into an integrated circuit was only a minor bump/improvement in a long series of “Moore’s Law” improvements. It just made bigger, more complex computing possible. My first job (1973) was an internship working on ARPAnet hardware. That too was a big deal but the Internet I worked on just a few years later was really just refinement/rework of the packet switching idea pioneered by ARPAnet, something that would have never happened without active government support and control. All of the “Internet” was significant at the society level given that there was only telephone wire before it; but the technology was simply available to Cisco/3-Comm, et al to leverage thanks, not to capital, but to government direction and subsidy. AltaVista Search (I was employee 11) was a research experiment that turned into an accidental and surprise success because, like the transistor it made something else (obviously) possible. Google/Yahoo! just copied/stole the idea. And so on… There were significant technological turning points over those years but even quantum computing, if it ever becomes operationally practical, is itself only an incremental change, a bending of Moore’s Law a bit more, not something new that was not there before. In other words, the electronics/computer/software/systems industry is now mature just like GM in cars, Union Pacific in railroads, Boeing in aviation. From now on, there is just incremental change/improvement, stupid ideas promoted by greedy fools and con men leading to the demise of ossified mega-corporations, and the diluting of once mighty monopolies, both corporate and governmental, as (now obvious) good ideas spread around the world.

All the Tech Bro’s really want to do and be like daddy in the glory days but can’t — not unless they do real, creative work in something really new. And you can’t do real work “strategizing” in a conference room on Sand Hill Road or where ever they hang out these days.

-----------—

1. A faith statement is just that, a statement made based on faith, not researched data. This concept of intelligence is so absurd that some “cognitive scientists”, many of them neurobiologists with an oversized philosophical materialist bent, deny that “consciousness” even exists simply because they cannot make sense of it in a “rational” materialistic framework. Yes, they just define it out of existence… Really. Really. I about fell out of my chair when I read that. I wonder what I was thinking… It’s like trying to measure the current in a live power cord with a micrometer which, if the wire is bare will remind you that the wrong tool does not work as they intended. This does not make faith statements false or absurd, just possibly dangerous to the clueless. Faith, informed faith is a good thing but there is also such a thing as blind faith which is often fatally dangerous to its practitioners.

2. Sand Hill Road is the street just north of the Stanford “Farm” campus that runs from downtown Menlo Park to the I-280 Interstate and beyond to the luxury homes of the “movers and shakers” in the Santa Cruz mountains. The one story “ranch style” offices overlooking I-280 are/were the home of the Venture/Vulture Capitalists who pored all that money into Intel, Cisco, Google, Sun Microsystems et al. I don’t know if they are still there but they were when some actual, real work was being done thanks to their enthusiastic, if frequently ignorant largess.

[END]
---
[1] Url: https://www.dailykos.com/story/2025/1/13/2296760/-AI-as-a-Bogus-Business

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/