"The Illusion of Thinking"

[ruigrazina]

Advanced AI suffers ‘complete accuracy collapse’ in face of complex problems,...

the-illusion-of-thinking.pdf

“Apple researchers have found “fundamental limitations” in cutting-edge artificial intelligence models, in a paper raising doubts about the technology industry’s race to [develop ever more powerful systems.
Apple [said in a paper published at the weekend that large reasoning models (LRMs) – an advanced form of AI – faced a “complete accuracy collapse” when presented with highly complex problems.”

Well now, this is a surprise… that ‘terrifying’pace of AI development apparently hit a slight problem…
These AI “experts” are beginning to sound a bit like those fantastic “experts” back in 2020…

[bluenose]

“If at first you don’t succeed, try, try, again”. There’s no stopping these people.

[Scarmoge]

… See Hubert Dreyfus - (1) What Computers Can’t Do and (2) What Computers Still Can’t Do. 1972 and 1992 respectively. I think both are available gratis on Archive.org

… yep, It (the computer) doesn’t “know” anything. It can neither assert nor propose. It can not experience nor remember. It can not interpret. It can not, as Antonio Damasio would say, “Have the feeling of what is happening”. But, and most importantly, it can not go “outside” its database.

… and no matter how many dyads are strung together to produce whatever length of dyadic chain at whatever speed is achievable … it will always be the Illusion of Thought.

[neru]

All the more scarier if our politicians could be basing all strategy on AI spewing out nonsense.

[neru]

As the “AI” Tech craze continues at peak levels… many substantial startups are now being exposed as a complete fabrication. Instead of revolutionary technology, they rely on overseas, underpaid workers who impersonate chatbots and “artificial intelligence” systems in order to secure massive investor funding, only to come crashing down months or years later in spectacular fashion.

Even industry leading companies like Microsoft, and Amazon are subjected to this problem, with an overall space that is now fully embracing the concept of useless, undeserved mania.

While the overall concept of Artificial intelligence certainly carries risks, and significant innovation with it, the current industry norm is anything but valid.

AI = Actually Indians ... (And Fraud)

[Sandygirl]

Ai can never have free will, Ai always has to obey a set of rules.
And now we must follow the rules as the computer takes more and more control over everything we do. It’s changing our behavior, it’s changing our language and it’s taking our free will. It’s getting more intrusive. WWW is such an accurate description, once you’re trapped in the web “I agree” it’s hard to break free.
I would not like to know how many hours I’ve been frustrated/upset when dealing with computer glitches and trying to figure out what the problem was. I would bet its over a 100 hours over the past 40 years.

[vardas3]

I would add Federico Faggin’s book, Irreducible: Consciousness, Life, Computers, and Human Nature.

[David8pies]