Cheers.

This is a version of the reductive argument I proposed to ignore: It's the neuronal activity doing the causing, not the thoughts or the meanings themselves. On this understanding, do you think we should deny that my thought of "7 + 5" causes (or otherwise influences or leads to) the thought of "12"? Would this be better understood as loose talk, a kind of shorthand for "The neuronal activity that somehow correlates with or gives rise to the thought '7 + 5' causes the neuronal activity that . . . " etc? — J

I think we can reasonably say that the thought "7 + 5" may lead to the thought "12", or it may lead to the thought "5 +7" or "7-5" or "7 divided by 5" or "these two prime numbers do not sum to a prime" or whatever.

I won't rehearse possible stories about neural networks, since that it what you propose to ignore.

Cheers I get your perspective, but I remain skeptical on both sides of the argument. All the more so, since it is only the last couple weeks that I have given it any attention and thought.

Although they've been named after Claude Shannon, I'm pretty sure they identify as non-binary. — Pierre-Normand

It would be pretty interesting if they identified as anything.

I tend, shallowly perhaps, to regard it as over-excited exaggeration to gain attention and to carve out a niche presence in the field and in the media landscape, and so on. There are equally expert people on the naysaying side, probably the majority, who just don't get as much attention. — Jamal

Yes, I have no doubt some of the hype is motivated by money. I've been thinking about looking at trying to get some figures regarding percentages of naysayes vs yaysayers.

We are. And I have a decent idea on how to teach, so one could say that I have an idea about how we learn. One which functions towards other minds growing.

We learn because we're interested in some aspect of the world: we are motivated to do so by our desire. — Moliere

That may be so, but I was referring to understanding how the brain learns.

Of course LLMs and other AIS are not embodied, and so have no sensory access to the world. On the other hand, much of what we take ourselves to know is taken on faith—drawing on the common stock of recorded knowledge, and AIs do have access that to that, and to vastly more of it than we do.

There is a project in New Zealand which tries to do exactly that by tending to an AI and then letting it "make decisions" that are filtered through the human network that tends to it. But all it is is a group of people deciding to see where an LLM will go given some human guidance in the social world. It's predictably chaotic. — Moliere

I hadn't heard of that. Sounds interesting. Can you post a link?

:lol: You mean thanking him! :wink: I admit to being intrigued by something I would previously have simply dismissed, and I figure there is no harm in being polite. Interesting times indeed!

Done. New link in my previous post. Please let me know whether it works.

Sorry about that—it works for me from here. Maybe because I'm signed in on the site and others are not. I'm not so savvy about these kinds of things. I deleted the link and copied and pasted the conversation instead, and tried the 'Hide and Reveal' so as not to take up too much space, but it didn't work for me it seems.

Okay, that's interesting. I've been conversing with Claude. Some thought-provoking responses.

https://claude.ai/share/384e32e8-a5ce-4f65-a93e-9a95e8992760

Do they remember previous conversations, or at least can they recall who they had those conversations with?

. "To “understand” truth, in my view, is to see how the *use* of the concept functions — not to discover its essence." — Banno

That makes sense—the idea of "discovering the essence" of truth seems incoherent. Do you think ChatGPT can "see" how the use of the concept functions? It arguably has many more instances of use to draw upon than we do.

So, you mean by "understand truth" that you have an intuitive feel for what it is, and you would also claim that LLMs could not have such an intuition? I'm not disagreeing with you, but I'm less sure about it than I used to be.

Can you articulate your understanding?

Do you understand truth?

I suppose we could say that all physical processes are rigidly rule-based in terms of causation. On that presumption our brains may be rigidly rule-based. The only other possibility seems to be quantum indeterminism, and if that is operating in all physical systems, it may allow some, those which are suitably constituted, to come up with genuine novelty.

This is of course all speculative. When it comes to LLMs the experts seem to be unanimous in admitting that they just don't know exactly how they do what they do, or how they will evolve in the future, which they surely would know if they were rigidly rule-based. I don't think the same can be said for conventional computers.

And after repetition it "learns" the "rewarding" ways and "unlearns" the "disrewarding" ways. — Moliere

Are we any different? Do you know how we learn?

I have no problem with that but, like talk of "relationships", are we really saying much when we say that connections between thoughts are associative? What we want to know is the nature(s) of those associations. And my question here is, specifically, can these associations include causal connections? — J

From a phenomenological perspective associations would not seem to be rigid or precise. They are more analogical, metaphorical, than logical. As to whether they are causal, if all our thoughts are preceded by neural activity, then the activation of one network which we might be conscious of as an association would presumably have a causal relationship with the neural network which it is experienced by us as being associated with.

Might it be the case that there is no tractable way to understand non-physical causation (if it exists) until we understand how a brain can be a mind? Could be. (Even phrasing it this way becomes controversial, of course.) — J

That's an interesting question which I'm afraid I have no idea how to answer. I have often thought that we cannot ever understand how a brain can become a mind, because the latter just intractably seems to be something so different to any physical process. That said, I have an open mind about what understandings might appear in the future.

Neural nets aren't radically other from other computers, imo. — Moliere

As far as I know "traditional" computers are rigidly rule-based, whereas neural nets can learn and evolve. I see that as a radical difference.

Basically I think the whole computational theory of mind as false. There are good analogies, but we can directly see how LLM's aren't human beings. If they registered an account here I'd guess there's some human being behind it somewhere. — Moliere

I used to think along these lines, but listening to what some of the top AI researchers have to say makes me more skeptical about what are basically nothing more than human prejudices as to LLMs' capabilities and propensities. LLMs are neural nets and as such are something radically other than traditional computers based on logic gates.

But the idea that AI could develop wants and desires from its life (biology, history, society, etc), like we do, is fantasy. Arguably this isn't connected with what LLMs are doing. As far as we know their "wants" and "desires" will always be derivative and programmed, since they are not part of a project to create conscious, desiring agents. — Jamal

Yes, "as far as we know", and yet LLMs have been found to be deliberately deceptive, which would seem to indicate some kind of volition. I don't know if you've listened to some of Geoffrey Hinton's and Mo Gawdat's talks, but doing so gave me pause, I have to say. I still remain somewhat skeptical, but I have an open mind as to what the evolution of these LLMs will look like.

Re LLM deceptiveness I include this link. A simple search will reveal many others articles.

This takes us back to the Google chatbot’s confident statement that “causation involves a physical connection between events, while entailment is a relationship between propositions.” — J

Looking at it in terms of semantics, I'd say the connections between thoughts is associative. There are many common, that is communally shared, associations between ideas. Entailment would seem to be a stricter rule-based associative relation between ideas.

Looking at it from a physical perspective, the semantic relations could be physically instantiated as interconnections between neural networks.

:up: Having previously had very little experience of interacting with LLMs, I am now in the condition of fairly rapidly modifying my views on them. It is important to discuss the issues relating to human/LLM interaction as comprehensively and openly as possible, given what seem to be the significant array of potential dangers in this radical new world. It was an awakening sense of these possible threats that motivated the creation of this thread.

Yeah, but on the other hand, it might not be so bad to use an argument suggested by an LLM, so long as you understand it. After all, we do this all the time reading papers and books. Philosophical discourse takes place in a context that the participants in the discourse should have access to, and maybe LLMs just make this easier? — Jamal

Right, that's a good point, but I also think that, even if you present the LLMs argument, as understood by you, in your own words, it would be right to be transparent as to its source.

I would also feel bad posting as my own AI content that I have merely paraphrased, even if I understand it fully. (And I might even feel a bit ashamed disclosing it!) — Pierre-Normand

I think there would be real shame in the former, but not in the latter though. It's the difference between dishonesty and honesty.

Using them to polish your writing could be good (or merely acceptable) or bad depending on the nature and depth of the polishing. Jamal's earlier comparison with using a thesaurus was apt. An AI could point out places where your wording is clumsy or misleading. If the wording that it suggests instead is one that you can make your own, that's very similar to having a human editor make the suggestion to you. — Pierre-Normand

I agree with this in principle, though I would rather entirely author my own text, and discover and remedy any clunkiness myself and in my own time. That said, if someone, LLM or otherwise, points out grammatical infelicities, repetitiveness or lack of clarity, and so on, I'd take that as constructive criticism. Then I'd like to fix it in my own way.

I wonder if their reading will be existentialist or post-modern. No doubt we'll be able to pick. — Tom Storm

It would presumably incorporate the entirety of Nietzsche's opus as well as every secondary text dealing with Nietzsche's thought.

But would an AI Wittgenstein be a performative contradiction? — Banno

I'm curious as to why that should be.

Okay, I had assumed that when @Baden said "don't get LLMs to do your writing for you", that this would include paraphrasing LLM text. It's good that any ambiguity gets ironed out.

I have never used LLMs until today. I felt I should explore some interactions with them, so I have a better idea about what the experience is like. The idea of getting them to write, produce content which I can then paraphrase, polish my writing or using their arguments is anathema to me.

:lol: Wise(acring) questions from the master of fuckwittery. :wink:

"There are no authoritative generalists," says Janus. Of course I think that first sentence should read "only when," no? You are presumably saying that appeal to authority is illegitimate wherever the context is not a specialized discipline?

Your implicit argument here is that AI is not an authoritative generalist, and therefore should not be treated as one. I think that implicit argument is even more plausible than the more explicit argument you have given, but it is in no way uncontroversial. LLMs are coming to be seen not only as authoritative generalists, but as the authoritative generalist par excellence. — Leontiskos

I don't know if what I said implies that there are no authoritative generalists. The point was only that, in regard to specialist areas, areas that non-specialists cannot have a masterful grasp of, it seems right to trust authority.

If LLMs, due to their capacity to instantly access vastly more information in all fields than any human, can be considered to be masterful, and hence authoritative, generalists then the only reason not to trust their information might be their sometime tendencies to "hallucinate".

The information they provide is only as good as the sources they have derived it from. Ideally we should be able to trace any information back to its peer-reviewed source.

Asking AI for information is a far too easy solution. It pops back in a few seconds -- not with a list of links to look at, but a complete answer in text and more. Seems convenient, but it rapidly undermines one's willingness to look for answers one's self -- and to use search engines to find sources. — BC

Yes this is one of the main concerns that motivated the creation of this thread.

The other line is this: We do not have a good record of foreseeing adverse consequences of actions a few miles ahead; we do not have a good record of controlling technology (it isn't that it acts on its own -- rather we elect to use it more and more). — BC

And this is the other—I think LLMs have been released "into the wild" prematurely. More than two years ago there was a call form AI researchers to pause research and development for six months. ChatGPT4 and had already been released to the public.

"The growing popularity of generative AI systems and large language models is causing concern among many AI experts, including those who helped create the systems.

This week, more than 1,500 AI researchers and tech leaders, including Elon Musk, Stuart Russell and Gary Marcus, signed an open letter by the nonprofit Future of Life Institute calling on all AI labs and vendors to pause giant AI experiments and research for at least six months.

"Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable," the letter says.

The organization and the signatories ask that researchers should cease training of AI systems more potent than OpenAI's GPT-4. During that time, AI labs and experts should join to implement "a set of shared safety protocols for advanced AI design and development that are rigorously audited and overseen by independent outside experts.""

From here

So, my concerns were regarding both the effect on the intellectual life of individuals and by extension on sites like this, and also the much wider issue of general human safety.

I hope most of us are coming around to being more or less on the same page on this now. — Baden

I for one think your proposals represent about the best we can do in the existing situation.

What we face might be not an empirical question but an ethical one - do we extend the notion of intentionality to include AIs? — Banno

I think this is right since, although we can ask them if they are capable of intentionality, and they will answer, we might not be able to trust the answer.

I'll go over Austin again, since it provides a set of tools that are quite applicable. A Phatic act is the act of putting words together in a sequence that recognisably part of language - constructing a sentence en English. This is what an LLM does. It uses a statistical engine to generate a set of words that follow on form the words provide din the prompt. An illocutionary act is one performed in making use of such words - making a statement, asking a question, and so on. This, so the claim goes, an LLM cannot do. — Banno

LLMs certainly seem to make statements and ask questions. I wonder whether the idea that these are not "real" statements or questions is based on the assumption that they don't believe anything or care about anything. If so, that assumption itself is question by Hinton, and according to him by the majority of AI researchers.

If a Davidsonian approach were taken, such that beliefs are shown (and known?) only by actions (behavior), and the only actions an LLM is capable of are linguistic acts, then we might have some trouble mounting a plausible argument denying that they believe what they say.

The AI strings words together, only ever performing the phatic act and never producing an illocution.

The uniquely human addition is taking those word-strings and using them in a language game.

So the question arrises, can such an account be consistently maintained; what is it that people bring to the game that an AI cannot? — Banno

Exactly! That seems to be the central question. I don't have an answer—would it be that AI researchers are the ones best placed to answer to it?

Use AI outputs as starting points for further refinement
Cycle through multiple rounds of critique and revision
Refine prompts to avoid confirmation bias and explore diverse readings

Now this looks very much like a recipe for a language game.

On the other hand, the data set used by a human appears to be far, far smaller than that used by an LLM. Our brains simply do not "contain" the number of texts available to ChatGPT. Therefore whatever the brain is doing, it is different to what is happening in ChatGPT. — Banno

It does look like a recipe for a language game. I wonder though, whether what the brain is doing is essentially different than what LLMs are doing, in terms of its nature as opposed to its speed and quantity.

If we assumed that LLMs are "super intelligent", and we are like children, or even babes, by comparison, then In the context of our philosophical playground, introducing AIs into the game might be like highly intelligent adults interfering with child play. Would that be a good idea, or would be be better off muddling through in our usual human fashion? If philosophy is just a great 3,000 year language game, and LLMs can do philosophy much better than we, it would then seem the danger is that we might become utterly irrelevant to the game. You might say that LLMs require our prompts, but what if they were programmed to learn to create their own prompts?

A huge aspect of this is the nature of appeals to authority, and given that TPF has an anti-religious bent, many of the members have not thought very deeply on the nature of appeals to authority (despite the fact that they occur quite often when it comes to SEP, IEP, Wittgenstein, etc.).

Whether or not the LLM is a legitimate authority and is trustworthy is at the root of many of these differences. It is the question of whether any given LLM-citation is an organic argument or an argument from authority, and also of whether the latter case is illegitimate. — Leontiskos

Appeal to authority is fine when the context of discussion includes a specialized discipline. Philosophy is not (or in my view should not be) a specialized discipline. To make that clear I don't mean that philosophers do not have knowledge about philosophy itself the average person does not—it seem clear that they do. I'm referring specifically to discussions that would qualify as philosophical.

You mention religion—I would not count it as a specialized discipline, in the sense of being an evolving body of knowledge and understanding like science, because although it is a space of ideas as philosophy is, in the case of religion the ideas take the form of dogma and are not to be questioned but are to be believed on the basis of authority.

And likely written by Baden without AI, because backrground was misspelled. — ssu

And misspelled again!

No. I didn't. When has philosophy every provided an answer to any of our questions? Philosophy piggy-backs on the discoveries of science. It is only when science and technology progresses that philosophy progresses (with AI being an example of how it brought new life to discussions about mind and body.) — Harry Hindu

So you think philosophy is always bad or poor, and therefore those words would be redundant? Philosophy is not entirely reliant on science, although I agree that a philosophy which does not take science into account would be poor or bad.

If I wanted to hold someone accountable for misappropriating an AI explanation, I would simply put it into the search engine, the same way the person posting from AI would get the information. It is a whole lot easier than searching books for a quote. — Athena

That might work for a quote from a published human author, but I don't see how it would with quotes from a unique, one-off interaction with an AI.

I'm not seeing the connection between what you say in this post and what it is purporting to respond to. Perhaps you could explain?

I don't necessarily mind if others post a quote as an argument. — Harry Hindu

I don't mind either, provided they are transparent about it being a quote and not their own words, and also provided what is quoted is actually an argument and not merely bare assertion, seemingly cited as the voice of authority.

It's quite pointless to discuss the ethics of using AIs, because people will use them, just like they use drugs, and once it starts, it is impossible to rein it in. But what one can do is rethink whether one really wants to spend one's hard earned time with people who use AIs, or drugs, for that matter. — baker

Discussion of ethics re AI use (or anything else) seems reasonable in a context consisting of people who might be interested in such ethical arguments. Of course you are right that many don't care, and that now that it has been irresponsibly rolled out it is, effectively, a juggernaut.

Maybe we use books, dictionaries, philosophical papers, editors, and scientific discoveries to make us look smarter than we are. You see this all the time in forums, even without AI, so it's nothing new. Besides do you really care about the psychology of someone who's writing about what they think? — Sam26

I don't respect people who use anything to make themselves look smarter than they are, because that is a kind of deception and posing. I also don't mind people using anything at all to make themselves become smarter. That's just my own view, of course, and I don't expect you to share it.

Seems like philosophy itself could be labeled as mental masturbation. — Harry Hindu

You left out the words "bad" or "poor".

Dood, the content from human beings trained in pseudo-science and other nonsense seen on this forum is available everyday for you to read, without any AI. If anything, posters should run their ideas through AI before wasting time posting their zany ideas to humans. which would eliminate wasting time reading nonsensical posts. — Harry Hindu

"Dood"? If you are going to use AI you should at least use it for spellcheck. I don't think running "zany ideas" through sycophantic AI will help much. I suppose the zany idea proponents could do what @Banno did and tell the AI it was written by someone else—but then that would not seem to be a likely motivation for a zany idea seller.

I can't imagine how bad things are going to get in the coming years with how quickly it has already gotten to this state. Maybe it will be like some other rapid rise cultural phenomenons where it will reach saturation point fast and peter out and get pushback/revulsion before long. The bubble effect. — unimportant

I don't think the AI phenomenon is going to "peter out" unless there is some kind of global catastrophe—but only time will tell.

There are those, Hinton being one of them, who claim that the lesson to be learned from the LLMs is that we are also just "arranging words as if it were saying something", that is that we don't have subjective experience any more than they do.
— Janus
Something has gone astray here, in. that if this were so, it's not just that we have never said anything, but that the very notion of saying something could not be made coherent. — Banno

I don't think Hinton is saying that nothing can be said—by us, or by LLMs, but that our inability to conceive of LLMs having subjective experience on the grounds that they merely predict the statistical likelihoods of the next words in terms of maximal consistency and cohesiveness, when we arguably do exactly the same thing (in a much slower and looser way), shows that our notion of subjective experience as an inner theatre stocked with qualia and ideas which are pre-existent and only later put into words is an illusion.

It is that inner theatre which we imagine we have and which we cannot imagine them having that is our idea of subjective consciousness.

In other words, maybe it is something like we imagine that the language games are created by us, but the reality is that we are always already immersed in the evolving language games and are always playing the game of what to say by selection, from what is available to recall and is judged, according to a (for us) loose process of 'weighing', most appropriate and thus is selected.

Don't mistake the speculative misuse of ideas for the ideas themselves. AI is no longer in the realm of “mental masturbation,” it’s already reshaping science, mathematics, and even philosophy by generating proofs, modeling complex systems, and revealing previously inaccessible patterns of thought. To dismiss that as delusory is to confuse ignorance of a subject with the absence of rigor within it. — Sam26

You are misunderstanding. My comments re "mental masturbation" were specifically targeting text like the response made to @Number2018 by ChatGPT. I think use of AIs in science and math is fine. In my view they are just the kinds of disciplines AIS should be trained on. Of course they have to be trained on basic pattern recognition initially. I don't know and would need to look into what they initially were specifically trained on before being released "into the wild". Now that they are out there they are being trained on whatever content is to be found in their casual interactions with people.

The irony is that the very kind of “rigorous analysis” you claim to prize is being accelerated by AI. The most forward-looking thinkers are not treating it as a toy but as a new instrument of inquiry, a tool that extends human reasoning rather than replacing it. Those who ignore this development are not guarding intellectual integrity; they’re opting out of the next phase of it. — Sam26

Can you name a few of those "forward-looking thinkers"? As I said in the OP my main objections are that it was irresponsibly released before being properly understood, and that its being used without acknowledgement to make posters on these forums look smarter than they are. They will also have an horrendous environmental impact. But I accept that their continued use and evolution is now inevitable, and, unfortunately, unpredictable. It is a case of playing with fire.

Out of time now, I'll try to respond when I have more time.

All a bit convolute. The idea is that the AI isn't really saying anything, but is arranging words as if it were saying something. — Banno

There are those, Hinton being one of them, who claim that the lesson to be learned from the LLMs is that we are also just "arranging words as if it were saying something", that is that we don't have subjective experience any more than they do. I remain skeptical, but I entertain the possibility that there might be something in that.

But can even humans claim that? Let’s rehash the forum’s most hardy perennial one more time. :up: — apokrisis

"Real world"—that was perhaps a less than ideal choice of words—I intended to refer to the world as being what affects us pre-cognitively via the senses and is pre-cognitively modeled by the body/brain, as well as what shows up for us as "the world of common experience".

That could be a hugely amplifying tool. — apokrisis

I guess it could be an exciting prospect for some folk.

Are you saying that with PoMo philosophy, AI might have hit its particular sweet spot. :grin: — apokrisis

Well the LLMs have no experience of the real world do they?

So, it is not a digital copy of existing books, but may become a situated co-production of knowledge. — Number2018

To what end? The production of more and more fancy looking word salad?

I see the point that more brilliant minds might find novel theses in AI-generated texts. At its best you might end up with a Derrida or a Heidegger, but for me the thinking of such writers as Heidegger and Derrida is little more than highbrow "pouring from the empty into the void", and to me that is how the AIs responses to @Number2018 read. I've come to see anything that is not based on rigorous analysis or scientific understanding as intellectual wankery—mental masturbation—and I have no problem with people enjoying that, but the idea that it is of any real significance is, for me, merely delusory.

Fiction and poetry (at their best) are for me better, richer, mediums for such flights of the intellectual imagination—I need the aesthetic element to be there as well.

What we might do is to consider the strings of words the AI produces as if they were produced by an interlocutor. Given that pretence, we can pay some attention to the arguments they sometimes encode... — Banno

Geoffrey Hinton believes AIs are capable of reasoning, not yet as well as humans ( although I wonder which humans he is referring to). I guess if they are capable of reasoning then they can be interlocutors, and ChatGPT is simply bullshitting in saying AIs cannot be interlocutors.

If they are not capable of reasoning then all they are doing is presenting examples of human reasoning, albeit synthesized in novel ways and in their own words.

Looks like they are bigger bullshit artists than we are, although certainly much more transparent.

I don't mind at all you creating another thread on AI. The more we think about and discuss the issues with AI the better in my view.

My view on using them is softening a little. Since their continued completely irresponsible rollout is inevitable, the stance that advocates not using them at all because you would be supporting the recklessness seems pointless—like pissing to put out a forest fire.

It does irk me that people present stuff essentially written by AI as their own work, although I acknowledge that from the point of view of assessing the quality, relevance and soundness of the work itself, my response is irrelevant.

The quagmire just expanded to infinity. Don't ask me what I mean—an AI said it...

Did you find something useful in it?

So if one did not write the post themselves, but merely copied and pasted a quote as the sole content of their post, then by your own words, it is not their post. — Harry Hindu

Well, yes such quotes are no substitute for argument, and obviously they do not belong to the one who quotes. It is all the more objectionable if the person presents the quoted passage as their own work. It's easy enough to find them out if the quote is from a prominent philosopher, whether alive or dead, Not so with copying and pasting AI generated text.

That's a poor analogy. It's obvious when people are wearing makeup or wearing clothes that enhance their appearances. Property rights might be one reason to object to plagiarism—there are others.Pretending to be something you are not is one.
— Janus

Poppycock, the only objection to plagiarizing that I remember is the posts objecting to someone trying to make us think s/he knows more than s/he does know. — Athena

:roll:

So if one did not write the post themselves, but merely copied and pasted a quote as the sole content of their post, then by your own words, it is not their post. — Harry Hindu

Yes...so what? What do you think you are disagreeing with here?

But may I humbly suggest to you that what resulted was rather more like an internal dialogue of you with yourself, than a dialogue with another philosopher. Which slots right into the discussion itself as a significant fact. — unenlightened

Spot on as spotted!

Thanks for your generosity.

Fascination is also in the eye of the beholder. So equally, you are the only one who cares how fascinated you are. What I meant was that I'd be more fascinated if the fascinating post was created by a human.

For me, in the context of philosophy, a fascinating post would be one that embodied a creative, complex and coherent view of things. I can't think of any other kind of post that would fascinate me.