I seem to be more solidly confident that so-called (post November 2022) AI is a dishonest and incompetent fake, than are most people.
Partly, this is because most people are on an AI-payroll (and I am not!) - but that doesn't explain my sureness.
The reason is I know that real science is dead in professional Big Science - and this has been the case for some decades.
What small functional progress we are seeing, is long-since Not real science, but engineering extrapolations of pre-existing science; plus a great deal of fakes and lies.
Real AI is impossible by my metaphysical assumptions (because real intelligence must be an attribute of a living Being. All the rest is tools, functional extensions, of Beings.)
But real AI is also impossible on the basis of actually-existing real science (which is now some decades old).
What has recently been touted as AI is (by its own admission) merely vastly-inefficient and -expensive, and top-down propagandized and coerced, engineering.
Which, to repeat, cannot be genuine AI.
And that is why I am so sure.
6 comments:
"Artificial intelligence" is a marketing name.
Though the definition of intelligence is a philosophical matter, we do know two things:
1. Computers are limited to computation
2. There are problems that can not be solved through computation alone (non computables). This has been logically proven. This limits what "AI" can do to computable problems
H. Indeed.
You can't replace intention with tricks.
Monkeys with typewriters won't write Hamlet, economic incentive structures won't make a country of scoundrels honest, social engineering doesn't replace religion, a universe doesn't come from a random scattering of particles, and you can't make intelligence from the statistical modelling of language.
@NLR "you can't make intelligence from the statistical modelling of language."
Exactly. It is just a psychological trick for getting computers to pass the Turing Test with unmotivated and inattentive people, under highly restrictive circumstances.
It has now been almost 3 years since the introduction of current "AI". Here are three predictions people have been making since the time it was first introduced:
1. It would lead to an explosion of creativity, particularly in the arts and indeed, would help to advance every field.
2. It would lead to the collapse of left-wing media and also make true information more widely available.
3. It would lead to economic prosperity overall, and also ordinary people would have new opportunities to make money.
How do those predictions look?
1. Rather than an explosion of creativity, there is more junk than ever before, much of it just to get money from unsuspecting people (for instance, I have seen obviously machine-produced math and science textbooks listed for sale). All fields have not advanced.
2. Media manipulation is going as strong as ever and true information is harder to find as search engines have become much worse.
3. Making money from "AI" is just a meme.
People who do so either already have dedicated followings, or are forced to use it for their jobs, or already have IT jobs, or are just out and out scammers. In other words, they already would have been making money, nothing special is going on. Also, the order of the day is to replace people so it has just led to more transfer of wealth from ordinary people to oligarchs and their minions.
The whole paradigm that people are using to making predictions is false, so naturally those predictions will fail. But we can see this empirically as well and it hasn't taken long.
@NLR - Well said.
It is one of the strange things about human beings, that most are indifferent to the failure of predictions - and do not regard failed predictions as evidence of a false theory.
Maybe it was the essence of real science, when it used to exist; that it was erected upon a process of making firm and specific predictions derived (as rigorously as possible) from its theories - then taking note and responding when the predictions failed?
Post a Comment