<!DOCTYPE html><html><head><title></title><style type="text/css">p.MsoNormal,p.MsoNoSpacing{margin:0}</style></head><body><div style="font-family:Arial;">My sympathies would be with your friend—until such time as a<b><i><u> "clone exactly like her ... behavior, words, or even existence..." </u></i></b>was demonstrated.<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">"Exactly" is a big word! and I would add "completely."<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">Even on a single dimension, say use of language, the standard of exact and complete is hard to satisfy.<br></div><div style="font-family:Arial;">I have no problem believing that a chat-bot could write an academic paper or either of my books; put together, and deliver in my voice, a lecture ; play bar-trivia at the pub; or carry on a convincing conversation. I have no doubt that, in the very near future, the same bot might be able to project a video that included mannerisms and simulation of the way I pace around a classroom.<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">But exactitude would require, not only, all the things I do do, and the idiosyncrasies in the way that I do them, but also the idiosyncrasies of my inabilities: I can never get the crossword clues involving popular culture, for example.<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">If a clone is built that "walks like a duck and quacks like a duck" but does not migrate or lay eggs; is it really a duck?<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">I would concede the equivalence issue of means or mechanisms behind the observable; e.g., it does not matter if the observed behavior results from electrons in gold wires or electrons in dendrites. But I would at least raise the question as to whether, in specific instances, a 'subjective' behind the behavior is or is not critical.<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">For example, and forgive the personal, you have mentioned being in pain all of your life. Would it be necessary for a bot to "feel pain" as you have in order to "act exactly like you?" Or is there an "algorithmic equivalent" possible for the bot to utilize in order to obtain unerring verisimilitude?<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">Then there is the whole question of experience in general. Would <b><u><i>I</i></u></b> really be <b><i><u>me</u></i></b>, sans the LSD trips over the years? If not, then how will the bot "calculate" for itself, identical or at least highly similar, experience equivalents.<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">Even if, in principle, it were possible to devise algorithms and programs that did result in behavior that mimicked Dave at every stage of its existence, will those algorithms be invented and programs written before the heat death of the universe? You cannot attempt to finesse this quest by invoking "self-learning" because then you need a training set that is at least as extensive as the 75 year training set that the mechanism you would have me be, has utilized to become me. <br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">I might agree that, in principle,<b><i><u> "A bot that acts indistinguishably from how you act *is* you," </u></i></b>I think the implication of the word "indistinguishably" is a bar that will never be attained.<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;"> davew<br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;"><br></div><div style="font-family:Arial;">On Tue, May 16, 2023, at 6:46 PM, glen wrote:<br></div><div style="font-family:Arial;">> That's a great point. To be honest, anyone who is accurately mimicked by <br></div><div style="font-family:Arial;">> a bot should be just fine with that mimicry, leveraging the word <br></div><div style="font-family:Arial;">> "accurate", of course. I mean, isn't that a sci-fi plot? Your bot <br></div><div style="font-family:Arial;">> responds to things so that you don't have to.<br></div><div style="font-family:Arial;">><br></div><div style="font-family:Arial;">> A friend of mine recently objected that "algorithms" are "reductive". I <br></div><div style="font-family:Arial;">> tried to argue that algorithms (in the modern sense of The Algorithm) <br></div><div style="font-family:Arial;">> can be either reductive or expansive (e.g. combinatorial explosion). But <br></div><div style="font-family:Arial;">> she was having none of it. I think her position boiled down to the idea <br></div><div style="font-family:Arial;">> that humans are complex, multi-faceted, deep creatures. And taking 1 or <br></div><div style="font-family:Arial;">> few measurements and then claiming that represents them in some space <br></div><div style="font-family:Arial;">> reduces the whole human to a low-dim vector.<br></div><div style="font-family:Arial;">><br></div><div style="font-family:Arial;">> So, for her, I can imagine even if she were cloned and her clone acted <br></div><div style="font-family:Arial;">> exactly like her, she would never accept that clone's behavior, words, <br></div><div style="font-family:Arial;">> or even existence as actually *being* her. There's some sense of agency <br></div><div style="font-family:Arial;">> or an inner world, or whatever, that accuracy becomes moot. It's the <br></div><div style="font-family:Arial;">> qualia that matter, the subjective sense of free will ... metaphysical <br></div><div style="font-family:Arial;">> nonsense.<br></div><div style="font-family:Arial;">><br></div><div style="font-family:Arial;">> A bot that acts indistinguishably from how you act *is* you. I guess I'm <br></div><div style="font-family:Arial;">> dangerously close to claiming that GPT-4 and Bard actually are <br></div><div style="font-family:Arial;">> sentient/conscious. *8^O<br></div><div style="font-family:Arial;">><br></div><div style="font-family:Arial;">> On 5/16/23 11:50, Marcus Daniels wrote:<br></div><div style="font-family:Arial;">>> I don’t really get it. Trump can go on a TV town hall and lie, and <br></div><div style="font-family:Arial;">>> those folks just lap it up. Sue a company for learning some fancy <br></div><div style="font-family:Arial;">>> patterns? Really? If someone made a generative model of, say, Glen’s <br></div><div style="font-family:Arial;">>> visual appearance and vocal mannerisms and gave him a shtick that didn’t <br></div><div style="font-family:Arial;">>> match up with his past remarks, I think I’d notice it right away. If <br></div><div style="font-family:Arial;">>> a GPT-X could fake Eric Smith, I can safely take the blue pill. Some <br></div><div style="font-family:Arial;">>> of our transactions will probably require more cryptographic signing. <br></div><div style="font-family:Arial;">>> Fine, they probably should have already.<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> *From:* Friam <<a href="mailto:friam-bounces@redfish.com">friam-bounces@redfish.com</a>> *On Behalf Of *Steve Smith<br></div><div style="font-family:Arial;">>> *Sent:* Tuesday, May 16, 2023 11:33 AM<br></div><div style="font-family:Arial;">>> *To:* <a href="mailto:friam@redfish.com">friam@redfish.com</a><br></div><div style="font-family:Arial;">>> *Subject:* Re: [FRIAM] Bard and Don Quixote<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> Jochen -<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> Very interesting framing... as a followup I took the converse <br></div><div style="font-family:Arial;">>> (inverse?) question To GPT4..<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> /If we consider an LLM (Large Language Model) as the Sancho Panza to<br></div><div style="font-family:Arial;">>> the Don Quixote of its human users, we can explore a couple of<br></div><div style="font-family:Arial;">>> potential aspects:/<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> 1. /Grounding and Practicality: Sancho Panza is known for his<br></div><div style="font-family:Arial;">>> grounded perspective and practicality, which complements Don<br></div><div style="font-family:Arial;">>> Quixote's idealism. Similarly, an LLM can provide factual<br></div><div style="font-family:Arial;">>> information, context, and practical insights to guide and<br></div><div style="font-family:Arial;">>> support human users in their interactions. It can offer a wealth<br></div><div style="font-family:Arial;">>> of knowledge and help bridge gaps in understanding, serving as a<br></div><div style="font-family:Arial;">>> pragmatic companion./<br></div><div style="font-family:Arial;">>> 2. /Balancing Imagination and Realism: Don Quixote embodies<br></div><div style="font-family:Arial;">>> imagination and creative thinking, while Sancho Panza offers a<br></div><div style="font-family:Arial;">>> more realistic viewpoint. Similarly, an LLM can enhance human<br></div><div style="font-family:Arial;">>> creativity by generating imaginative ideas and expanding<br></div><div style="font-family:Arial;">>> possibilities. At the same time, it can bring rationality and<br></div><div style="font-family:Arial;">>> critical thinking, ensuring that the generated content aligns<br></div><div style="font-family:Arial;">>> with facts and reason./<br></div><div style="font-family:Arial;">>> 3. /Providing Different Perspectives: Sancho Panza often offers<br></div><div style="font-family:Arial;">>> alternative viewpoints and challenges Don Quixote's romanticized<br></div><div style="font-family:Arial;">>> notions. Similarly, an LLM can present diverse perspectives,<br></div><div style="font-family:Arial;">>> consider different angles, and engage in nuanced discussions<br></div><div style="font-family:Arial;">>> with human users. It can help users explore varying opinions and<br></div><div style="font-family:Arial;">>> broaden their understanding of complex topics./<br></div><div style="font-family:Arial;">>> 4. /Ethical Guidance: Sancho Panza acts as a moral compass,<br></div><div style="font-family:Arial;">>> advising Don Quixote on the right course of action. In a similar<br></div><div style="font-family:Arial;">>> vein, an LLM can provide ethical guidance by highlighting<br></div><div style="font-family:Arial;">>> potential biases, promoting inclusivity, and encouraging<br></div><div style="font-family:Arial;">>> responsible use of language and information. It can contribute<br></div><div style="font-family:Arial;">>> to creating a more informed and ethical digital environment./<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> /However, it's important to note that an LLM, as a machine learning<br></div><div style="font-family:Arial;">>> model, lacks human experiences, emotions, and deeper understanding.<br></div><div style="font-family:Arial;">>> While it can offer valuable assistance, it may not fully grasp the<br></div><div style="font-family:Arial;">>> complexities of human interactions or possess the wisdom and<br></div><div style="font-family:Arial;">>> empathetic qualities of a human companion like Sancho Panza.<br></div><div style="font-family:Arial;">>> Therefore, while an LLM can serve as a useful tool, it should be<br></div><div style="font-family:Arial;">>> complemented by human judgment and critical thinking to navigate the<br></div><div style="font-family:Arial;">>> nuances of real-life situations./<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> I am a couple of hours behind on the live feed Senate hearing on AI <br></div><div style="font-family:Arial;">>> <<a href="https://www.judiciary.senate.gov/committee-activity/hearings/oversight-of-ai-rules-for-artificial-intelligence">https://www.judiciary.senate.gov/committee-activity/hearings/oversight-of-ai-rules-for-artificial-intelligence</a>> listening in fits and starts between other things:<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> 1. I was definitely impressed with Altman (OpenAI), Montgomery (IBM)<br></div><div style="font-family:Arial;">>> and Marcus' (NYU) thoughtful and extrapolative responses rather than<br></div><div style="font-family:Arial;">>> perhaps the more usual defensive/deflective/adversarial style that<br></div><div style="font-family:Arial;">>> these hearings often have... I don't know who chose this<br></div><div style="font-family:Arial;">>> particular lineup but I thought both Montgomery and Marcus made a<br></div><div style="font-family:Arial;">>> good complement to Altman. If Google and Microsoft and ??? had been<br></div><div style="font-family:Arial;">>> there it might have reflected more "competitive" or "showy" answers?<br></div><div style="font-family:Arial;">>> 2. I was impressed with the Senators (compared to my fairly low<br></div><div style="font-family:Arial;">>> expectations). Even Kennedy and Hawley managed not to do their<br></div><div style="font-family:Arial;">>> usual grandstanding and and snide sniping. Altman handed Hawley's<br></div><div style="font-family:Arial;">>> question "why don't we just let people sue you?" (as a mode of<br></div><div style="font-family:Arial;">>> regulation/oversight) back to him quite deftly (in the style of "ass<br></div><div style="font-family:Arial;">>> with both hands") by responding simply "I know of no reason people<br></div><div style="font-family:Arial;">>> CAN'T sue us today if we cause harm". Marcus chipped in pretty well<br></div><div style="font-family:Arial;">>> outlining how the current laws that *might* apply are not well<br></div><div style="font-family:Arial;">>> suited for many reasons.<br></div><div style="font-family:Arial;">>> 3. I felt that all three witnesses walked the fine line on the question<br></div><div style="font-family:Arial;">>> of a "moratoriam" fairly deftly, acknowledging that they endorse the<br></div><div style="font-family:Arial;">>> spirit of not running headlong and thoughtlessly into the future but<br></div><div style="font-family:Arial;">>> at the same time there is no obvious practical way to implement and<br></div><div style="font-family:Arial;">>> enforce this, but that they are all enforcing their own limits on<br></div><div style="font-family:Arial;">>> how fast (and thoughtlessly?) they might roll out development to the<br></div><div style="font-family:Arial;">>> public...<br></div><div style="font-family:Arial;">>> 4. In closing Blumenthal suggested having ChatGPT rewrite the lyrics<br></div><div style="font-family:Arial;">>> to "Don't Stop" (thinking about tomorrow (McVie-Fleetwood Mac) which<br></div><div style="font-family:Arial;">>> I took to heart. I was not impressed with it's results and won't<br></div><div style="font-family:Arial;">>> burden the list with it. I'm guessing Blumenthal did *not* actually<br></div><div style="font-family:Arial;">>> do that but like Quixote, simply saw the windmill and thought it<br></div><div style="font-family:Arial;">>> might be a giant?<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> On 5/16/23 11:30 AM, Jochen Fromm wrote:<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> I have asked Bard ( bard.google.com) today about Don Quixote from<br></div><div style="font-family:Arial;">>> Cervantes, and if a large language model would be similar to a Don<br></div><div style="font-family:Arial;">>> Quixote without a Sancho Panza.<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> Here is what Bard replied:<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> "In a way, large language models can be seen as Don Quixotes without<br></div><div style="font-family:Arial;">>> Sancho Panzas. They are trained on massive amounts of text data, and<br></div><div style="font-family:Arial;">>> they are able to generate text, translate languages, write different<br></div><div style="font-family:Arial;">>> kinds of creative content, and answer your questions in an<br></div><div style="font-family:Arial;">>> informative way. However, they are not always able to distinguish<br></div><div style="font-family:Arial;">>> between reality and fiction, and they can sometimes make mistakes or<br></div><div style="font-family:Arial;">>> generate text that is not accurate or helpful.<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> Sancho Panza, on the other hand, is a realist who helps Don Quixote<br></div><div style="font-family:Arial;">>> to stay grounded in reality. He provides Don Quixote with common<br></div><div style="font-family:Arial;">>> sense and practical advice, and he helps him to avoid making mistakes.<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> Large language models could benefit from having a Sancho Panza<br></div><div style="font-family:Arial;">>> figure to help them to stay grounded in reality and to avoid making<br></div><div style="font-family:Arial;">>> mistakes. This could be done by having a human being or another<br></div><div style="font-family:Arial;">>> large language model to review the text that the large language<br></div><div style="font-family:Arial;">>> model generates. This would help to ensure that the text is<br></div><div style="font-family:Arial;">>> accurate, helpful, and appropriate"<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">>> -J.<br></div><div style="font-family:Arial;">>> <br></div><div style="font-family:Arial;">><br></div><div style="font-family:Arial;">> -. --- - / ...- .- .-.. .. -.. / -- --- .-. ... . / -.-. --- -.. .<br></div><div style="font-family:Arial;">> FRIAM Applied Complexity Group listserv<br></div><div style="font-family:Arial;">> Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom <br></div><div style="font-family:Arial;">> <a href="https://bit.ly/virtualfriam">https://bit.ly/virtualfriam</a><br></div><div style="font-family:Arial;">> to (un)subscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br></div><div style="font-family:Arial;">> FRIAM-COMIC <a href="http://friam-comic.blogspot.com/">http://friam-comic.blogspot.com/</a><br></div><div style="font-family:Arial;">> archives: 5/2017 thru present <br></div><div style="font-family:Arial;">> <a href="https://redfish.com/pipermail/friam_redfish.com/">https://redfish.com/pipermail/friam_redfish.com/</a><br></div><div style="font-family:Arial;">> 1/2003 thru 6/2021 <a href="http://friam.383.s1.nabble.com/">http://friam.383.s1.nabble.com/</a><br></div></body></html>