Jack, Jon, List Jack, I read your post as expressing what I understand to be a classical AI view: the machine performs formal symbolic operations only, while the human supplies mind. I've been taking a shot in this thread at adding a little nuance to that view. For even if the machine lacks mind in the full sense (which it most certainly does), the signs it generates may enter a larger semeiotic process in which genuine inquiry is developing. Now that's a different claim from saying the machine itself is intelligent, and it's also different from reducing the machine to 'merely' a formal symbol system.
Further, it had seems to me that perhaps Peirce offered the philosophical foundation for this view: Since thought itself proceeds through signs, any technology that fundamentally alters our ability to generate and manipulate signs alters the conditions under which inquiry develops. Consider this snippet which Peter Skagestad quotes in part in the paper pointed to in a related thread. [T]he logical machines that have thus far been devised can deal with but a limited number of different letters [i.e., in algebra]. The unaided mind is also limited in this as in other respects; but the mind working with a pencil and plenty of paper has no such limitation (*Writings*, Vol. 6, p. 71). This is to say that, for Peirce analysis and such is no longer the isolated mind but the *mind-in-relation-to-an-external symbolic system*. I would suggest that today's AI systems represent another step in the development of symbolic systems. Unlike books, diagrams, etc., they are, as I've frequently noted, *responsive *symbolic systems which can *generate new signs in response to other signs*. They can summarize, compare, reformulate, and even suggest hypotheses.. In other words, the novelty of contemporary AI may be the emergence of what may prove to be a new kind of symbolic system, namely, one capable of participating actively in the ongoing semiosis of an inquiry. As I've put it before, AI appears to me to potentially alter the 'ecology of inquiry'. It changes not just the speed with which we manipulate symbols (which, of course, is remarkable in itself), but the kinds of semiotic interactions that become possible. That, I think, is precisely why Engelbart's *Intelligence Augmentation* and Skagestad's interpretation of Peirce in light of IA fit together rather naturally as they direct attention away from intelligence as an isolated property of a machine and toward the evolving symbolic systems through which inquiry itself grows. I see you making two claims: 1. AI consists of formal symbolic manipulations and is not a mind or conscious subject. 2. The human user supplies meanings and purposes while the AI system merely carries out formal operations on symbols. That position does indeed seem quite close to what Jon has been arguing: Genuine semiosis, intelligence, and inquiry remain properties of the larger triadic process grounded in human agency and continuous 3ns. I'm asking whether formal symbol manipulation *exhausts* what is happening. Peirce, I believe, would distinguish between a. the physical or formal manipulation of signs and b. the semiosis in which those signs become interpretants within an ongoing inquiry. Indeed, that distinction has been at the center of my exchange with Jon. So to sum up: As I see it, today's LLMs are not classical formal symbol systems in that they do not manipulate explicit logical symbols according to hand-coded rules. Instead, they operate through learned symbols (and other representamen) distributed across huge neural networks. Nevertheless, one can yet argue, as Jon has, that LLMs' internal operations remain fundamentally discrete computational processes rather than the continuous triadic semiosis that Peirce associated with intelligent action. OK. I can agree with that. And I agree with Jon that the *most* one could characterize them as is *quasi-intelligent,* although I am coming to believe that even that may be claiming too much. All I am now arguing is that the symbols that LLMs genera te may enter a larger semiotic process in which genuine inquiry develops. That's a different claim from saying the machine itself is a mind; but it's also different from reducing the machine to 'merely' a formal symbol system. Best, Gary R On Fri, Jul 10, 2026 at 5:19 PM Jack Cody <[email protected]> wrote: > List, > > AI exists within F:SS (filter: symbolic system). Organic "intelligence" or > "being" (more broad) is beyond that, always, whatever it is, also whilst > within/using such. > > It can never, as far as I can tell, bridge that gap and for good Peircean > reasons regarding the most basic definition of what is symbolic (agreed > upon meaning — also, I'd add, meaning which can be agreed or disagreed > upon, which seems to me not a deviation at all). Anyway, AI is built from > F:SS where people merely use it. There is no mind to it. Don't see how you > get around that or why you'd want to. > > Best, > Jack > ------------------------------ > *From:* [email protected] <[email protected]> on > behalf of Jon Alan Schmidt <[email protected]> > *Sent:* Thursday, July 2, 2026 11:49 PM > *To:* Peirce-L <[email protected]> > *Subject:* Re: [PEIRCE-L] A Peircean Argument for the Reality of AI > Intelligence > > Gary R., List: > > I agree that Pietarinen's suggestion of an "overlapping triadic relation" > whose correlates are the utterer, the sign, and the interpreter is quite > interesting. However, I am not yet convinced that it is an accurate > conceptualization, especially since communication is not the only > manifestation of semiosis. According to Peirce, the utterer and interpreter > are *dispensable*; the object is the essential ingredient of the utterer, > and the interpretant is the essential ingredient of the interpreter (EP > 2:403-10, 1907). There are signs that have no utterer--"symptoms of > disease, signs of weather, etc." (EP 2:496, 1909 Feb 26)--and signs that > have no *actual *interpreter, i.e., no *dynamical *interpretant. > > Even so, Peirce does describe a sign as "a medium of communication" and "a > determination of a quasi-mind" (EP 2:390-2, 1906), where what is being > communicated is "a Form" and "it is necessary that there should be two, > if not three, *quasi-minds*" involved (EP 2:544n22, 1906). Specifically, > "signs > require at least two Quasi-minds; a *Quasi-utterer* and a > *Quasi-interpreter*; and although these two are at one (*i.e.*, are one mind) > in the sign itself, they must nevertheless be distinct. In the Sign they > are, so to say, *welded*" (CP 4.551, 1906). Moreover, "The quasi-mind is > itself a sign, a determinable sign" (SS 195,1906 Mar 9); and "the > *Communicational > Interpretant* ... is a determination of that mind into which the minds of > utterer and interpreter have to be fused in order that any communication > should take place. This mind may be called the *commens*" (SS 196-7, EP > 2:478, 1906 Mar 9). > > Accordingly, I suggest that it is better to avoid referring to > utterer-sign-interpreter as an "overlapping triadic relation" and instead > identify utterer, commens, and interpreter as the three quasi-minds > involved in any event of semiosis--when a dynamical object, the essential > ingredient of the utterer, determines a sign-token, the essential > ingredient of the commens, to determine a dynamical interpretant, the > essential ingredient of the interpreter. In this account, can "AI" systems > participate in semiosis by serving as such quasi-minds? If so, is that > sufficient to ascribe genuine intelligence to them? As someone who views > the entire universe as one immense sign, such that discrete things and > their dyadic reactions--including digital processing operations--are > degenerate outcomes of continuous and triadic semiosis, my current answers > are yes to the first question and no to the second. > > Frankly, I dispute the assertion that "AI" systems are "able to respond, > criticize, reformulate, compare alternatives, and extend an inquiry"; I > still maintain that they merely *simulate *these behaviors. They do not > possess *agency*, the ability to *choose *whether and how to perform such > tasks; they only do exactly what they are told by humans, in exactly the > way that they have been programmed to do them by humans, just like other > computer programs. They produce an admittedly very sophisticated *illusion > *of collaboration by mimicking human language, including the use of first > person, but they are really just additional tools facilitating inquiry by > humans. Their output consists of degenerate dyadic reactions, not genuine > triadic interpretants. > > Regards, > > Jon Alan Schmidt - Olathe, Kansas, USA > Structural Engineer, Synechist Philosopher, Lutheran Christian > www.LinkedIn.com/in/JonAlanSchmidt / twitter.com/JonAlanSchmidt > > On Wed, Jul 1, 2026 at 11:02 PM Gary Richmond <[email protected]> > wrote: > > Jon, List, > > Pietarinen's paper has provoked a thought that I had not previously fully > appreciated and which seems relevant in response to your critique of my > thesis that AI represents a form of intelligence. What especially struck > me is not just Pietarinen's regarding nonhuman sign users as > potentially viable participants in inquiry, but his suggestion that *Peirce's > account of communication may itself involve two overlapping triadic > relations*. One, the *semiotic triad, Object - Sign - Interpretant*, > which describes semiosis as such: the object determining the sign, > determining an interpretant which may itself become a new sign in the > continuing growth of inquiry. > > Pietarinen also identifies a second triad, what might be calls the > *communicative > triad, **Utterer - Sign - Interpreter*. The self-same sign that mediates > between object and interpretant also mediates between participants in > inquiry. Communication, in this view, is not something added onto semiosis, > rather, *semiosis is intrinsically communicative*. Peirce's definition > that a sign is "something which stands to somebody for something" perhaps > deserves to be taken quite literally in human communication, perhaps > especially that communication involved in inquiry. > > If this is correct then inquiry begins to look quite different. It is not > just a series of individual, personal acts of reasoning; and neither is it > simply transmission of information between inquirers. Rather, it is > possible to see *inquiry as a developing network of overlapping sign > relations in which interpretants generated in one communicative exchange > can become signs in another*. Thus, the *community of inquiry *is > nourished by the continuing semiosic processes through which meanings are > created, criticized, revised, and further developed. > > In my view this also helps clarify the place of AI within inquiry. In an > earlier post I remarked that, as I see it, *AI can participate in > semiosis without possessing mind*. Rather, Pietarinen's analysis suggests > that *AI systems can function as a participant within the communicative > dimension of semiosis.* From my own experience, I believe that one can > say that the kind of continuity that develops in an extended human-AI > interaction isn't at all that involving an artificial personality, but *merely > the continuation of the inquiry*. > Jon, you raise what seems to me to be an important qualification. You ask > whether present or future AI systems might actually possess ends of their > own, or, do they merely function as sophisticated means -- tools -- by > which we humans pursue our ends. At present I strongly agree with you that > AI does not exhibit autonomous purposiveness in the sense Peirce associates > with organisms, individual humans, and communities. However, my proposal > does not require such autonomous purposiveness. > > In my view, participation in semiosis needn't imply independence of > purpose. For example, scholarly books and academic papers certainly preserve > and transmit signs without possessing ends of their own. AI differs from > them, however, in *actually** being able to respond, criticize, > reformulate, compare alternatives, and extend an inquiry *(btw, my sense > is that one can't really see this capacity without at least experimenting > with AI in an inquiry of her own). So for me the question is not whether > AI is or can become an autonomous inquirer, but *if responsive AI sign > systems can contribute genuine interpretants to inquiries* whose ultimate > aims, however, remain those of human individuals and communities. > > Whether this ultimately proves to be a plausible way of understanding > contemporary AI remains an open question. Still, it strikes me as a > distinctly Peircean question. If communication is recognized as intrinsic > to semiosis rather than external to it, then the possibility that > responsive AI sign systems can participate in the communal growth of > inquiry no longer seems conceptually outlandish. At the very least, I > believe that it is a philosophical question well worth investigating. > > I would certainly be eager to read the thoughts of especially young > interested people as this question regarding AI seems to me now to be a > kind of philosophical "Wild West" (as one young educator whom I'll discuss > in a future post, put it). > > Best, > > Gary R. > > _ _ _ _ _ _ _ _ _ _ > ► PEIRCE-L subscribers: Click on "Reply List" or "Reply All" to REPLY ON > PEIRCE-L to this message. PEIRCE-L posts should go to [email protected] > . > ► <a href="mailto:[email protected]";>UNSUBSCRIBE FROM > PEIRCE-L</a> . But, if your subscribed email account is not your default > email account, then go to > https://list.iu.edu/sympa/signoff/peirce-l . > ► PEIRCE-L is owned by THE PEIRCE GROUP; moderated by Gary Richmond; and > co-managed by him and Ben Udell. >
_ _ _ _ _ _ _ _ _ _ ► PEIRCE-L subscribers: Click on "Reply List" or "Reply All" to REPLY ON PEIRCE-L to this message. PEIRCE-L posts should go to [email protected] . ► <a href="mailto:[email protected]";>UNSUBSCRIBE FROM PEIRCE-L</a> . But, if your subscribed email account is not your default email account, then go to https://list.iu.edu/sympa/signoff/peirce-l . ► PEIRCE-L is owned by THE PEIRCE GROUP; moderated by Gary Richmond; and co-managed by him and Ben Udell.
