I’ve found an appropriate if simple summation of “Artificial Intelligence.” It could be a conceptual hurdle, though, for the true believers of AI: “Artificial Intelligence” is just a machine. A digital kind, but mechanical in operation. It has no more “intelligence” than a can-opener.
However, it can be trained, which is a step up from even the best can-openers. This leads some to project upon it the answer to their longings for whatever they feel might make a better world. And then some use the idea to make money.
I read recently that AI is best described as a text extrusion device, and this tickled me, for I’ve played with ChatGPT a bit in that capacity. I’m writing novels and a number of times I’ve played with concepts and characters, bouncing them off the Large Language Model. Typically I’ve asked for lists based upon my prompts. Then I inspect them for angles I haven’t thought of, or creative combinations.
Reading tea leaves
Occasionally I might get a text that was the seed of a good idea extruding out of the platitudinous mediocrity of most of the responses. But even this was mostly me being sparked creatively. It was as if I were telling someone’s fortune by reading tea leaves, and a chance configuration reminded me of a person or lit up a memory to expand upon. It wasn’t like the tea leaves were intelligent.
(I’m not really sure if tea leaves are read like that but they should be!)
I would never ask ChatGPT to write anything for me. I haven’t used it for awhile now, and the models move on. But the tone, the voice of the machine revealed by the text was so bland and insensate. They don’t have bodies, you know.
Although if you want ChatGPT to write old-fashioned rhyming verse, it does that well.
I feel I can detect the overall “style,” if we may call it that, of AI extrusion more easily now.
There are some limited things that so-called AI can probably do better than humans, especially in the calculating parts of technical fields like medicine or engineering. But even there, they couldn’t be trusted for their judgment, for example, where lives matter. They would have to “know” what was at stake, and how could they? They are like intricate clockwork, yet knowing nothing of time.
Health aspects
And it may not be good for your health, to have too much reliance on AI. Here are two references that give pause: 1) An article on Forbes, Your Brain On AI: ‘Atrophied And Unprepared’ citing a recent study, and 2) an article by Daniel Dillu on Medium, How Over-Reliance on AI Could Lead to Cognitive Atrophy, back in 2023, which analyzes how this might occur.
The first article points out the problem for individuals, in which people decrease their critical skills in favor of AI, but also warns of “mechanized convergence,” in which reliance on AI among many leads to less diverse and creative outcomes.
A recent post in Time magazine, based upon a Lancet medical study, notes that doctors who regularly used AI actually became less skilled within months.
The second article in Medium above looks at how the negative effects could develop. For example, what happens to our innate ability to remember when we place all responsibility for that on a machine? Memory is like a muscle, it needs to be exercised.
One of my science fiction ideas, once the internet got established and before AI was so prominent, was that everyone put their memories on the Web, and heaven help us when the internet goes down. This had its similarities to Isaac Asimov’s story about people unable to calculate by hand, due to reliance on calculators, and how shocking it was when someone re-invented that.
Financial bubble
The financial bubble represented by the investment of billions and billions in AI is starting to show a few leaks since AI hasn’t resulted in a real game changer. Investor confidence could be shaken! Not to speak of the huge potential environmental impact, especially the AI need for fossil fuel or nuclear energy and impact on the use of water which are crucial issues everywhere in these times of climate change.
A recent study found that 95% of the attempts to use AI in business have failed to make money, and the remaining 5% are mostly about its efficiency in creating vast volumes of spam for annoying marketing purposes.
But the real interest for me is what the AI bubble reveals about human nature. Or maybe it’s just about the sad lack of connection for many people in our culture.
The meaning of ELIZA
Joseph Weizenbaum in 1966 created a highly scripted chatbot called ELIZA which applied pattern matching rules to converse with people.
That some embraced it as their private psychotherapist was shocking to the computer scientist. More than a few people seriously thought it was responding with empathy, but in reality ELIZA merely reflected back what a person asked of it.
And ELIZA only had 200 lines of code. “What I had not realized is that extremely short exposures to a relatively simple computer program could induce powerful delusional thinking in quite normal people,” Weizenbaum wrote.
Now imagine those minimal lines of code expanded to include the entire internet as a pool to parrot from.
There’s an article on IEEE Spectrum that describes the phenomenon of ELIZA, “Why People Demanded Privacy to Confide in the World’s First Chatbot.” It discusses this tendency to attribute empathic understanding to simulcra like ELIZA. And of course ChatGPT and the others.
Weizenbaum found, the article says, that “the simulation of intelligence, rather than intelligence itself, was enough to fool people.”
After ELIZA, Weizenbaum’s mission became “to remind people that their machines were not as smart as they were often said to be. And that even though it sometimes appeared as though they could talk, they were never really listening.”
A cautionary tale.
The most recent videos of this kind that I’ve enjoyed are those from Amy Shafer, a harpist and classical pianist who calls herself Virgin Rock. She does present in her musical reaction videos an image like a restrained music teacher or librarian. But she breaks into enthusiasm at different aspects of the rock music she is, apparently, discovering for the first time. All while wearing pearls.
Another I recently watched is her appreciation of
The best presenters to me are not making squeals of delight or putting on exaggerated expressions, but it is the quality of the listening that comes through, even if they don’t make a big deal of their reaction.
Recent Comments