Threaded Index     Date Index     BC



Re: I often don't/cannot respond to posts...

Posted by Manstuprator on 2024-May-2 00:25:16, Thursday
In reply to The relevance of your links alludes me posted by diogenes on 2024-May-1 10:43:20, Wednesday




... due to "health issues" (Issues my ass! They are PROBLEMS not ISSUES! God damned "newspeak"!)

Either I am "a person with" (more "newspeak") "long covid" or I am dying from one of several fatal degenerative neurological disorders (I am "life span" challenged, you could say...)

I just finished making a post on the main board that gave some essential security advice to a new poster, and just making that post has almost completely depleted whatever neurotransmitters I have left that remain available in my brain.

Oh, well, I'll try to respond to your post.

First, I'd like to say that I find your posts interesting and informative, and you seem much more knowledgeable on many topics than I am! And though I can understand the terminology that you use, it is often only in my passive and not in my active vocabulary.

Anyway, I am not trying to fight with you, but I would just like clarify some things that I believe I may know a bit more (pun intended) about than you do regarding how computer programs work.



I have a confession. The thought of AI makes me unutterably depressed.

I don't think it merits getting depressed over, but it certainly should be concerning. It doesn't depress me, but it does worry me...

Take my sigpic. In its full glory (it's slightly blurred compared to the picture I originally submitted to BC)...

Your original sigpic image must have been larger than around 10,000 bytes, which is the standard size limit for allowable sigpic images, so one of the cogs "compressed" it (reduced the image quality) to meet the limit. If you edit your sigpic again, ensuring that it is smaller than 10,000 bytes (you could simply reduce the dimensions, but keep the high quality) then we could see it's full glory. ;-)

... it was a pretty good simulacrum of Van Gogh. Yes, something seems wrong with the knuckles on the right hand, but overall, it's pretty good. Possibly as good as an actual Van Gogh, and with a subject matter I find more appealing.

Yes, it's quite good, and the subject matter IS "more appealing"!

But this didn't take days for a painter to paint. It took seconds for an image generator.

Yes, that's amazing!

And it seems there is no area - not the creative arts, not scientific discovery - in which AI will not eventually be able to do things better than a human.

OK -- this is where I take issue with you. I have read a lot about AI, and actually I think the acronym is a misnomer -- it should be SI (Simulated or Synthetic Intelligence).

When I think of artificial/synthetic things, what are brought to mind are things that actually are equivalent to the natural things they are intended to substitute for. Artificial sweetener does actually taste sweet, though it is created by technological "artifice". Artificial/synthetic Vitamin C molecules are indistinguishable from the naturally occurring ones, as are many other chemicals synthesized -- but identical -- to the naturally occurring ones.

But "Artificial Intelligence" is NOT truly intelligent. Computer programs are linear (though these days there are CPUs that can do parallel processing of the coded software) and just follow (blindly) a series of steps determined in advance by the programmer.

I am not current on IT, but the last time that I checked, there didn't exist a computer with even the brain power of an ant.

I linked to those articles because I agreed, due to my understanding of the concepts involved, with the "assertions" they made.

Correct me if I am wrong, but in your post you have made unsubstantiated assertions just as they did in their articles.

And then what? Humans will never again be able to feel pride over performing a useful task for their community, for nothing they can do can help their community as well as an AI can if set the same task.

I strongly disagree with you here. When you say "set the same task" you are assuming that a programmer has written (or could even write) code that can encompass all the variables/possibilities/requirements involved with that task.

A computer is just a dumb machine. It only does what it is instructed to do (that's why computer programs are also referred to as "computer instructions":
https://www.capterra.com/glossary/instruction/

Without those instructions, a computer is just a fancy collection of high-tech components, with nothing to do.

In fact, humans will no longer do anything at all. We shall become mere consumers.

In the West, most of us perform service jobs, anyway, and don't really produce any real products at all. At the same time, we are consumers. Industry is mostly mechanized (which includes "dumb" robots, which are just as happy to crush the operator against the wall as to install the rear panel on an automobile chassis. So much for "intelligence"...).

Will we be happy? Should we be happy? It is this that makes AI depressing. Humans are surplus to requirements.

AI does NOT make us surplus. AI only simulates intelligence, but is not in itself intelligent. Sure, in certain situations it can fool us, but no AI can pass a truly comprehensive Turing test.
"Theoretical physicist and science popularizer Michio Kaku summed up and compared current A.I. to cockroaches:
“At the present time, our most advanced robots, some of which are built in Japan and also at MIT, have the collective intelligence and wisdom of a cockroach; a mentally challenged cockroach; a lobotomized, mentally challenged cockroach.”
... an interesting and mostly accurate (I believe) article:
https://bigthink.com/technology-innovation/what-animals-is-ai-currently-smarter-than/

Even as consumers, probably machines will soon outstrip us. There will be AIs that can appreciate works of art in the way that an expert can, and thus on a much deeper level than most of us. We will become inept even as consumers.

Were what you say true, then AI should be able to write guaranteed hit songs, hit movie scripts, and the finest emotionally engaging art. I believe the likelihood of those things happening are small (though perhaps "pop" music songs could be created that would sell well). Human emotions, I believe, are something that (except for rudimentary simulations) will always be beyond the reach of AI.

The only solution I can think of is that human communities will come to reject AI. But the human race in general cannot reject AI whilst it is divided into mutually hostile territorial states. The only possibility is that once self-aware artificial superintelligence takes over everything, it will dismantle all existing states as posing a threat to life and to itself.

The solution that I can think of is that the smarter folks will realize the threats that relying on AI involve, given its demonstrable shortcomings and failings. We won't let AI have control of things that could lead to our destruction. But, then again, we have "smart" systems in place that control power grids, but which have not been "smart" enough to prevent cascading failures that black out entire regions, even nations...

I have faith (misplaced?) that we are not so stupid (or, at least, we have members of our species who are not so stupid) as to let disaster occur.

Well, ideas anyone? How can we not be depressed at the coming age of complete human redundancy?

We will never be made redundant, I believe.

But I could be wrong...

M.
I've left out a lot in the above. But that's it! My neurons are shot! I need an AI program to continue posting for the rest of the evening!




Follow ups:



Post a response:

Nickname:

Password:

Email (optional):
Subject:


Message:


Link URL (optional):

Link Title (optional):


Add your sigpic?