The danger from #AI and #ML does not come from the #technology, but its owners and controllers. The Luddites didn't smash the looms because they were anti-tech (contrary to modern representations[1]), but because they were forced out of income[2]. That the #tech #oligarchs can't be trusted is on display time again, currently with the #mathematics community[3]. #society #history #politics #economics #epsteinclass [1]economicshelp.org/blog/6717/ec… [2]bloodinthemachine.com/p/unders… [3]wired.com/story/openai-is-piss…
Similar posts
3 posts close in meaning, closest first
The #LLM is something different from #AI, in my admittedly jaundiced opinion. In fact I would say that it's OBVIOUSLY less than intelligent not merely because it's stochastically mimicking training data without applying abstract reasoning, but for a reason far more obvious: it requires intelligent human beings to make decisions for it. This has been one of the many scandals that GenAI boosters and apologists have been attempting to suppress with trumped-up stories of LLM achievements: cbsnews.com/news/ai-work-kenya… time.com/6247678/openai-chatgp… and many more.
Show more of this post
If #OpenAI et alii have created an unprecedented "artificial general intelligence" then why would sweatshop workers need to be making key decisions on its behalf? Furthermore, consider that one of the biggest and most reliable excuses one gets from those devoted to normalizing their dependence upon LLM sludge is: "I always double check it, I'm very careful." Now that's almost certainly a self-delusional lie: #tech people tend to have very poor self-awareness about what they're actually doing and they habitually tell whatever self-flattering lies support their illusions of superior intelligence and unique wisdom. They're aware they need to pretend to be responsible users of LLM sludge so...they pretend, even if their real goal is to crank out code or text automagically without bothering to look at it (for maximal "efficiency".) So they lie. But they feel they MUST tell this lie: it's as close as they can come to admitting that stochastic LLM glop is, in fact, stochastic glop which should never be trusted without thorough review—a process that basically erases whatever gains in "efficiency" and "velocity" are sought from #GenAI in the first place. So this too is an indirect admission that LLMs simply aren't intelligent. They require someone with ACTUAL intelligence to sort through the sludge and correct it.
Nevertheless it seems also clear that the #LLM is mimicking more than just bits and pieces of its training materials. One is taken with the sense that LLMs and the whole "Generative #AI" phenomenon are built to mimic how the typical #tech geek tends to solve problems--their intellectual approach to, well, everything basically, an intellectual approach practically devoid of intellect because its ultimate goal is to do away with thinking altogether. Thinking is time-consuming, emotionally expensive, and thus bad for #efficiency in the John Carmack or Elon Musk sense.
Show more of this post
The elite techlords don't want to do ANY of it; they want to think that correct decisions and infallible snap-assessments are just gushing out of them all the time, like they're wired directly into the Truth of All Things...like they're a prophet in other words. Ultimately this fallacy has a religious nature: the high-tech "gurus" (remember, they called themselves that appropriated religious word) really have become like a priesthood of Technology[tm], technology turned into an artificial magic system like in Isaac Asimov's Foundation. Which, you know, they probably think of as aspirational. If "thinking" is the word. It's this quality of instantaneous snap answers that has been impressed onto all the stochastic #GenAI thingummies. Of necessity the techlords are compelled to claim that their thing is (almost) supremely intelligent to a degree impossible for the puny human meatbrain to comprehend...because that's the only way to support the premise that the LLM isn't merely emitting trash but unparalleled wisdom. (more)
My closing point is just this: the #tech sector has backed itself into a corner. They are the acme and exemplar of this entire abusive culture, this social system in which there's supposedly only a few specially gifted geniuses who must drag the sheeplike masses along (by trickery and forcible compulsion if necessary) to the glorious future waiting for "Western Civilization", so long as it doesn't get any funny ideas about disobeying corporate wishes.
Show more of this post
The technology sector sits atop this authoritarian corporatist juggernaut, claiming credit for having made it all possible, because software is weaved into every last corner and crevice of this corporate behemoth. And the techlords--so few of them when you really think about it--feel they've got the entire show right in their grasp. Either they get unlimited license and resources to build their Machine God, or they crash the entire system: that's the #OpenAI Sam Altman threat, more or less. And what does it alllll depend on? What's the secret sauce, already regarded and advertised as utterly and totally necessary to the corporate tech sphere? LLM goo. Easily corruptible LLM goo. The glue that holds the U.S. #economy together. ~Chara of Pnictogen