November 20, 2023

There are a lot of frauds in so-called artificial intelligence becoming visible at the rationality and social levels. Even calling the process intelligence is fraud, as important elements of intelligence cannot be built into computers, such as acquiring knowledge and applying it to questions.

How should computers define knowledge, when humans take centuries to refine knowledge? Most modern knowledge is science based. No one is claiming computers are going to evaluate science journals to determine what knowledge is. Most of physics is in error.

One of the largest, over-riding frauds of so-called artificial intelligence is in the nature of the process. The new-fangled form of artificial intelligence scans through huge amounts of information and extracts some sort of average for each question. Quasi averaging is not originating something new‌ in a significant manner.

Also limiting in that process is that there needs to be a large amount of information available in digital form to extract the quasi average from. That's why the promotional material focusses on literature. Where there are a lot of words, a lot of extracting becomes possible. But outside of literature, the information base rapidly becomes harder to find.

Another place where there is a lot of data to extract from is imagery. So fake images are contrived to look like some average representation. What value is there in doing that? Misrepresenting someone is about all that is being done with the fake imagery.

Which gets to a major fraud on the subject. Why is the stuff being programmed to do that? It's only going to do what it is programmed to do. It didn't need to be programmed to fake someone's imagery. But faking imagery is one of the few things the junk can do; so the exploiters designed something that gets used for perpetrating fraud in misrepresenting imagery instead of finding something more constructive to do with the junk.

Finding something more useful gets back to what computers were always doing, which is speeding up processes that would take humans a long time to do, such as mathematics. Increasing precision is another long-term project of computers, as automotive technology is using to put parts in place. That stuff is called automation.

Why add a quasi-averaging process to what computers do? Simply because it is there. No real value is visible in doing that. Yet the junk is promoted as if it were unlimited in solving human problems.