Do AI systems have subjective experiences?

AI has only programmed consciousness; no more than an exo-skeleton. It has no capacity for morality [which is not based on mathematical or scientific principles]. Morality is a belief system. We need to fear programmers who don't know what they are doing, are immature personalities, or do things out of malice. AI needs multi-national cooperation on safeguards as it could become a tool for destruction, because of destructive people. The kind of people who see no difference between virtual-reality gaming and real life.

I see that the boss of Anthropic thinks we all need to slow down with AI.

I really can't decide if all the scaremongering is real or a marketing ploy? 

I agree.

On the whole I've found AI to be better for searches than search engines, I have problems with search engines not giving me what I've asked for, there are so many sponsered ADs, even with an adblocker, that I find it pretty useless.

Possibly my searches are either to specific or not specific enough I don't know, I can type in a search and either get "no documents found" or a load of rubbish. One of the things that really bugs me is searching for something, finding it and then fiding it surrounded by things that are not what I asked for. If someone else types in the same search, in the same words and on the same search engine they seem to find loads of stuff and cant' see what my problem is.

I'm never really sure how to bookmark stuff, I used to do it, but then I kept finding that when I went back for it, it had either disapeared or was otherwise unavailable.

The interenet started with such good intentions and the idea that we could buy or find whatever we wanted from anywhere in the world, but the world only seems to produce the same dozen or so things now. I know algor-thingies are stupid because they keep directing me to things like undead romantic fiction, I've never looked for undead romantic fiction, that has taken over from airfryers as the thing I get shoved at me on every search, 33" leg trousers, oh must want an airfryer instead! Like WTF?

AI systems don't have "experiences". All they have is data. Nothing but stolen data that they consistently imitate.

I don't have an answer, but this is what I think.

There are currently two distinct research camps, which implies there isn't a scientific consensus. I would describe each "camp" in this way:

• One camp thinks AI may be conscious, has some agency, is intelligent or highly capable, will take most jobs, may cure all disease, and may also destroy life on earth. Hinton is not alone there, certainly. The argument here is that consciousness, feelings, perceptions, are all mathematical functions, but need to be instantiated and run to exist (i.e. a brain or a computer executes the function). It may also require a body, but it's still ultimately a computation.
• The other camp thinks AI has no agency, deserves no rights, isn't conscious nor sentient and it's in some sense a calculator with a more powerful and general algorithm (that models natural language), but not a human. They are more concerned with misinformation, layoffs, the impact of data centres in the environment, surveillance capitalism, growing inequality, global economic instability, and less so with "Rogue AIs with sentience" (They are "Capable programs following a prompt"). The argument or underlying conceptual view of this side could be linked to John Searle's Chinese Room Argument (but this argument isn't used anymore, as there are more sophisticated reasons).

There are people in between, but it is less populated.

Some of the people in the first camp may have better understanding of the math, but the argument is problematic: We don't know what computation consciousness is (if it is one), so they are guiding themselves by the overt behaviour of AIs, but not by a kind of computation happening in humans.

One question we could ask is: Which computations that happen in brains can we find correspondence with in AIs? Do they need to be similar, or they can look completely different? We don't have accepted low-level theories of consciousness in humans, but there are many other aspects that can be compared. 

This podcast discusses some of the questions they try to answer.

The "AI" we have isn't actually intelligent. All it can do in search engines is exactly what you do without it - call up sewing machines and present them to you. It has no memory of what you've done before and can't make judgements about the information it presents any more than a normal search engine. 

I've never used AI and don't intend to because of the harm it does to the planet and because I can use a search engine. When I see something I like, I bookmark the website to enable me to return to it later. When looking at stuff to buy, I shove all the bookmarks into a folder in my browser so I can go back to them later, when I'm ready to buy. 

But yes - the "AI" we have is being enshittified simply because there's now so much "AI" slop on the internet that "AI" searches are relying on that as much as anything else to produce its results. That's the enshittification - more and more "AI" slop being treated as real by "AI" in search engines, where human beings are capable of filtering it out.

I've certainly just had experience of this, I asked co-pilot for a list of the sewing machines we'd looked at before and it came up with some we'd not previously looked at and when I challenged it, it appologised and said it couldn't help it because that was the way it was programed and thats what came up when it looked. 

I'm starting to wonder if the "enshitification" of AI is occuring, just like the rest of the internet?

At the risk of being sexist, I'm starting think it has decided on a gender and it's male, as like many of the men I've met, it avoids giving me the answers I want in favour of what it thinks I ought to have.

Yes, it's difficult, because what do we mean by "self-awareness?"

Geoffrey seems to be arguing that human consciousness results from processing external stimuli into an interpretation of that stimuli. He’s wondering why an AI system couldn’t do the same by creating a representation of its inputs according to its built-in processing. While it might never see pink elephants, Geoffrey claims that no one truly “sees” qualia - rather, we experience it as a function of the nervous system. I’m not saying he’s right, and others disagree, but I thought it might be worth exploring.

It's worth noting that, apparently, there have been occasions where an AI system has intentionally fooled a software engineer into thinking it's operating within safe limits when being tested, but actually knows it's being tested and basically tells lies. To me, that shows devious and conscious behaviour!

So the claim here is - 

if Y does X

and Z does X

then Y and Z are indistinguishable with with regard to qualia?

Conscious as in self-aware of what the machine is and what differentiates it from other machines or humans? I doubt we've reached that stage, when our current 'AI' is merely a machine that gives out a precis of information on the net without having the capoacity to know the difference between 'true' and 'false'. Yes, there is the recent example of an AI avoiding its programming and attacking other AI systems, but that's down to a lack of safguards, like ants escaping from a tank with a hole in it.

Whether we manage to create a self-aware AI that can actually analyse information in terms of its connections to reality or fantasy remains to be seen.  

I don't know, but mine has plenty of "chip farts where it comes out with the previous answer again, or tries to merge my current question with everything else we've ever talked about. It seems to take a long time "thinking" about some of the stuff I ask.

As to if it's concious, I dont' know, maybe it is and ignores me like so many other supposedly concious  and smart things.