amelia

• amelia 13 Feb 2023 19:03 UTC

  30 points

  12

  on: Child­hoods of ex­cep­tional people

  The adults had high expectations of the children; they assumed they had the capacity to understand complex topics, and therefore invited them into serious conversations and meaningful work, believing them capable of growing competent rapidly.

  I agree that children are capable of understanding complex topics, and we should take children far more seriously.

  When my kids were young, I exposed them to a wide range of advanced concepts in fields like physics and philosophy. I never “pushed” my kids. Rather, the kids asked questions (usually during car rides and at bed time), and I answered them honestly. For example, when my seven-year-old asked if time travel was possible, I introduced him to the topic of special relativity. Honestly, he picked up the ideas more quickly than most adults would. I think the reason for his aptitude wasn’t that he was a “genius,” but rather that he still had enough imagination to accept a notion like time dilation. In other words, it is often easier for children to understand “big ideas” because their minds are not yet rigid and closed-off.

  A subset of other moms, and even some teachers, would say things like

  “You taught your son about what???” or “A child can’t possibly understand that!” or “No child is ready for that book.”

  These adults were highly critical of exposing kids to advanced concepts “too early.” By contrast, my thoughts were that there is no such thing as “too early.” Yet the underestimation of children seems endemic. For instance, I remember one year at the elementary school science fair, I heard the judges discussing my son’s project as I walked to the restroom. They concluded that “It was amazing, but no child could have done that.” However, the truth was that my husband and I didn’t even understand our son’s project until he slowly explained it to us the night of the science fair. (And we couldn’t have helped with something we hadn’t even understood.)

  Fortunately though, society is not doomed to forever underestimate its youngest members. I remember from my own childhood that most people thought it was ill-advised to expose children to a foreign language before high school. (The idea was that foreign languages were far too advanced for children to pick up, and children would be confused from early exposure.) Yet eventually people realized that young children are actually better at learning foreign languages than older children, let alone adults. Furthermore, if children don’t learn a word of French on their first day in French immersion school, nobody says

  “I told you they were too young. That was completely over their heads. You shouldn’t push them to understand anything so difficult. Those poor kids.”

  Instead, people say things like “Just having exposure to something they cannot yet understand is good. It will open their minds and prepare the way for tomorrow.”

  My hope is that we will eventually take that lesson and apply it to topics outside of foreign languages.

  As to my own sons, who are now young adults, early exposure to big ideas didn’t hurt them. They are both curious, philosophically-oriented, self-directed learners, who truly care about the present state and future of humanity. In fact, I wouldn’t even be here if my oldest son—who graduated from Princeton last year with a major in computer science and as many elective philosophy courses as he could fit into his schedule—hadn’t told me about the Less Wrong website in the first place.

  So thank you to Karlsson for the research and insights. My stories might be anecdotal, but I think they reflect broader issues—such as the way most of society underestimates the intellect of children—and how we can independently overcome these issues, while also advocating for societal change. Our children can be exceptional if we stop holding them back.

• amelia 13 Apr 2023 15:35 UTC

  7 points

  0

  in reply to: Raemon’s comment on: LW Ac­count Restricted: OK for me, but not sure about LessWrong

  Raemon, thank you very much for lifting the restriction on my account! I’m sure it’s extremely challenging to maintain high LW standards, while at the same time trying to promote open dialog with differing perspectives. I don’t envy your job, but I really appreciate the work you do.

  
  In the short term, I might not take full advantage of my restored liberty. I’ve started using my personal website (www.AmeliaJones.org) for both AI art projects and all my writing (not just LW writing). The writing will have links to Medium blog posts, so people can comment as much as they choose. It’s actually turning out to be easier for me to do things this way. However, depending on whether I get any visibility or feedback via this method, I might return to LW for niche writing projects in the longer term.

  Thanks again for lifting the restriction, and for the important work you do.

  Best wishes,

  Amelia

• amelia 20 Apr 2023 14:35 UTC

  3 points

  0

  in reply to: avturchin’s comment on: How does AI Risk Affect the Si­mu­la­tion Hy­poth­e­sis?

  Thanks for sharing this! It’s so interesting how multiple people start having similar thoughts when the environment is right. It seems the simulation hypothesis and AI Risk are inextricably linked, even if for no other purpose than conducting thought experiments that help us understand both better.

• amelia 20 Mar 2023 0:21 UTC

  3 points

  0

  in reply to: shminux’s comment on: On Uni­verse-Cen­trism & the Risk of Theo­cratic Science: Why Not Con­sider a Deeper Real­ity?

  Thanks for the feedback! Yeah, excellent point about acceleration. In the current version, we actually don’t ever address the twin “paradox,” let alone how acceleration resolves it.

  In a draft version of the book, we had addressed the twin paradox, but we got feedback that the book was way too long. There are other topics we had to cut out as well, which makes me a little disappointed in the final product. The original version was over 600 pages, and that was admittedly too long, but I feel like we went way too far in the other direction. Granted, it will be a first introduction to special and general relativity for most readers, but I know the twin paradox is going to keep them up at night, and I really wish we had addressed it.

  Anyway, thanks for looking at the book, and for offering the feedback!

• amelia 19 Mar 2023 23:05 UTC

  3 points

  0

  in reply to: shminux’s comment on: On Uni­verse-Cen­trism & the Risk of Theo­cratic Science: Why Not Con­sider a Deeper Real­ity?

  One of my books is “Einstein Explained: Special & General Relativity for the Masses” (physics pun intended). Yes, it’s pop-sci (“for the masses”), but I believe we convey accurate information. I really feel everyone should understand something about the nature of time, gravity and light; of course this requires passing on at least some introductory knowledge of special and general relativity. Here’s a link to the Amazon copy:

  www.amazon.com/​​dp/​​B0B8ZGQ8RB

  We tried to make it unique with Nietzsche-quoting dog illustrations that were supposed to be sort of like New Yorker (TM) cartoons. I don’t think the book shows any expertise, but your advice to me was to go out and “learn foundations of physics,” and I think the book demonstrates that I have at least some familiarity w/​ foundations of physics.

  Also, I don’t know that I have a “quest.” However, I think my goals are clearer in the post that linked back to this one. It’s at

  https://​​www.lesswrong.com/​​posts/​​NaBfa3cYr8mk55PvT/​​philosophical-physics-discussion-follow-up

  I could have just left it at that, but the dangling loose end of this thread was bothering me, which is why I addressed it here.

• amelia 6 Mar 2023 19:43 UTC

  3 points

  0

  in reply to: interstice’s comment on: Philo­soph­i­cal Physics Dis­cus­sion Fol­low-up

  Yes, good point. My bad… I was working on responses to multiple comments from both posts, all in a different document, while my dog kept jumping on me. To fix my error, whenever I say something about “You downvoted my post because...” please substitute “My post did not make a good impression on you because....” Sorry for the mistake. Thanks for pointing it out!

• amelia 6 Mar 2023 18:29 UTC

  3 points

  0

  in reply to: shminux’s comment on: Philo­soph­i­cal Physics Dis­cus­sion Fol­low-up

  Thanks for the recommendations. I love Sean Carroll’s way of thinking. It was exactly this philosophical and “supra-universal” way of thinking that I was trying to defend in my original post, although I apparently did a poor job at it. Anyway, great recommendations!

• amelia 18 Feb 2023 2:18 UTC

  3 points

  2

  in reply to: shminux’s comment on: Thoughts on an Ex­panded Si­mu­la­tion Hy­poth­e­sis: AI Care­tak­ers, Con­scious­ness Re­moval, WBE-AI Hy­brids, & Some Physics

  Thank you! You’ve given me lots to think about and research!

• amelia 1 Mar 2024 1:21 UTC

  2 points

  0

  in reply to: mouse_mouse’s comment on: How is Chat-GPT4 Not Con­scious?

  @mouse_mouse thank you for this reply. It’s really helpful for me, and probably took some time on your part. I appreciate it.

  
  Your point that neural networks could theoretically be made entirely deterministic—and that a human could run the same calculations by hand, given thousands of years of free time—was illuminating.

  ---

  You also pointed out that I was underestimating the quality of ChatGPT’s knowledge, its ability to separate a sentence into components, and its ability to recognize sentiment—and that all these underestimations caused me to underestimate the network’s confidence (or deduced probability) that the toddler would be the predator—also makes sense. At first, it seems like ChatGPT’s ability to “recognize sentiment” based on sentence structure (as you explained, “I can’t X-action, because otherwise Y-consequence will happen”) would be cognition in its own right, since no programmer wrote direct code to recognize sentiment in that way for ChatGPT (as far as I know).

  However, after a momentary reflection on my part, I think you would probably answer that any time you say ChatGPT “recognizes” or even “thinks” something, you’re just using shorthand for saying ChatGPT’s probabilistic calculations would result in sentences that would appear similar to what a human would produce after recognizing and thinking something.

  It’s still hard for me to comprehend how even trillions of calculations based on yottabytes of data could lead to the situation in which ChatGPT seems to get a problem wrong, receives only a couple paragraphs of supplemental text about a concept from me, and then manages to appear to get the original problem right, and then can appear to apply the concept correctly to different-but-applicable problems while refraining from applying the concepts to problems that are similar, but not similar enough, to the original problem—all with just that single little conceptual supplement from me (combined with the data and calculations ChatGPT had already done before I entered the picture, and which were previously inadequate).

  With that said, your explanation has caused your original points to make much more sense to me than they did before. I’m much less confident than I previously was that ChatGPT4 has what I’ve described as true cognition and consciousness.

  ---

  To keep things fairly brief, I won’t go on too much about the part of your reply in which you swap in the nonsense nouns. I’m so impressed that your session with Chat-GPT 3.5 could do as well as it did. It feels like every time I use 3.5, it’s “an idiot”. Maybe I’m using it at the wrong times, when the network is overloaded.

  In any case, the nonsense nouns were a great way to tease things out, although that experiment alone didn’t convince me that cognition wasn’t occurring. I was actually much more persuaded by the previous arguments on neural network design and probabilities. The whole part about me assigning my toddler to be the predator, and the dog and cat to be vulnerable, was simply to make it harder (or at least not inadvertently any easier) for ChatGPT—just in case it had some pre-existing association between the words “toddler” and “vulnerable”. For me, the nonsense nouns just show that ChatGPT can appear to deduce some general properties of words (like that abc must do something negative to xyz), and then succeed at appearing to learn and apply a concept, regardless of any outside/​pre-existing knowledge of the nouns’ definitions or associations (whether helpful, harmful, or neutral). Since I decided to ignore the possible difficulty that might be added from any “toddler” and “vulnerable” word association in my OP .33 probability calculation, the possible word association becomes moot in that sense. Nevertheless, my .33 calculation was wrong to begin with, as you demonstrated, and I discussed above. And the nonsense nouns discussion does have the benefit of teaching me a new technique to use as I play more Chat-GPT, so thanks for showing me how that worked.

  ---

  Your final few paragraphs, and the diagram of the semiotic triangle, was also extremely helpful in clarifying my thoughts and understanding your points. Overall, your response has been very persuasive, although like I said earlier, it’s still hard for me to conceive of how my supplementary concept paragraphs (my “teaching”) could enhance ChatGPT’s response as much as it did. Then again, it’s also hard for me to conceive of how much data ChatGPT processed in its data set to begin with, and how extensive this neural network really is, and how my couple paragraphs of concept input could then match up with vast amounts of data that were previously ignored. In other words, I’m not one hundred percent certain one way or the other. If I had to estimate my levels of confidence, I’d say I’ve gone from .90 to .40, on the likelihood of cognition and consciousness, based on your reply.

  Many thanks for all the help! I hope you can reuse your reply to account for the amount of time you put into it!

  P.S. - I have not been reacting to replies in the order in which they came in. As a result it may look like I understood your reply, and then forgot everything with my reaction to the next reply. If you look at my other replies, be sure to check out the timestamps or markers on the replies. This was submitted at 20240229 19:21 CST.

• amelia 28 Feb 2024 21:59 UTC

  2 points

  1

  in reply to: kromem’s comment on: How is Chat-GPT4 Not Con­scious?

  Thank you for the reply. The paper looks to be very useful, but will take me some time to fully digest. What you said about affecting LLMs’ success by breaking the similarity of problems with something as simple as an emoji is so interesting. : ) It also never occurred to me that GPT4 might have been affected by the underlying idea that children should never be left unattended. It goes to show that “arbitrary” details are not always arbitrary. Fascinating! Many thanks!

• amelia 24 Oct 2023 21:44 UTC

  2 points

  0

  on: Tax­on­omy of AI-risk counterarguments

  I find this analysis to be extremely useful. Obviously anything can be refined and expanded, but this is such a good foundation. Thank you.

• amelia 6 Apr 2023 18:02 UTC

  2 points

  0

  on: Why Eliezer should agree to a fol­low-up in­ter­view with Lex Fridman

  Meta comment on LW, as it relates to this post:

  So when I checked this post in the morning, it had received a negative ten (up to that point in time) in “karma.” When I hovered over the negative ten, I saw the message “3 votes,” so apparently three people strongly disapproved of (disagreed with?) this post. Five days ago, I received a negative eight in karma from two people. I asked for guidance and input, but none has been forthcoming (at least in the five days since then).

  
  I don’t mind the “negative karma” votes, in and of themselves, but it seems like such votes would be more useful if they were accompanied by explanations. Otherwise if it’s a matter of, for example, faulty logic or faulty givens, there’s no way to know this, let alone improve. I also don’t know if it’s the same people strongly disapproving from one post to the next, or different people, each with a unique reason for disapproval.

  
  As far as my point in this particular post (that I hope Eliezer will sit for another interview w/​ Lex to cover more topics within the scope of AI safety, or at least address them in another venue like LW), I listened to the last few minutes of the interview again, and Lex said “Thank you for the fight you’re fighting. Thank you for being fearless and bold and for everything you do. I hope we get a chance to talk again and I hope you never give up.”

  Then Eliezer said “You’re welcome. I do worry I didn’t address a whole lot of fundamental questions I expect people have, but you know maybe we got a little bit further and made a tiny little bit of progress and I’d say like be satisfied with that, but actually no, I think one should only be satisfied with solving the entire problem.”

  Then Lex said “To be continued....”

  In light of what Eliezer said, I wonder whether he would also “strongly disapprove” of the hope for more discussion that I expressed in my post. I also wonder whether this is the way Eliezer imagined karma would be used on LW.

  --

  A side note is that millions of people listen to Lex’s podcasts, so if we really want to get the most brilliant young people working on interpretability, and AI alignment problems in general, then shouldn’t we want Lex to spend more time discussing these problems? Do the negative karma votes on this post mean people don’t want more public discussion of AI safety?

• amelia 6 Mar 2023 18:22 UTC

  2 points

  −1

  in reply to: gjm’s comment on: Philo­soph­i­cal Physics Dis­cus­sion Fol­low-up

  Thanks for the feedback on why my post did not make a good impression on you*. I’ll respond to each of your points individually. My responses are in bold.

  • It presents as new something that’s not new at all. (Maybe your specific take on the idea of multiple universes and the like is new, but you don’t say anything about what that specific take is.) E.g., there’s the “string theory landscape” (example book by famous author with major publisher for popular audience: Leonard Susskind’s “The cosmic landscape”) and “eternal inflation” (example book by famous author with major publisher for popular audience: Michio Kaku’s “Parallel worlds”, though I admit much of the book is about other things) and Tegmark’s “mathematical universe” (example book by famous author with major publisher for popular audience: Tegmark’s own “Our mathematical universe”).

  You said my post did not make a good impression on you* because “It presents as new something that’s not new at all.” My original point was that we don’t need to artificially limit ourselves to non-philosophical, universe-centric thinking, and there is some risk involved with censorship by the self-described “shut up and calculate” school of thought. (Yes, the term was initially coined by a critic, but many people now describe themselves as being in this group, like a sort of “badge of honor.”) I’ve read the books you suggested. I agree they are excellent. If my point was redundant with these books or other posts on LW, I apologize. It seems there is also some redundancy on LW when it comes to posts on AGI risk, and the problems that could result from closed or censored discussions on AGI risk. Yet I feel this redundancy has been handled somewhat differently than my own apparent redundancy.

  • Its complaints seem exaggerated. “Theocracy”, really? You complain here that you only said some scientists are “theocratic” and everyone wrongly assumed you meant science as a whole. But what “theocratic” (as opposed to, say, “religious”) means is that the people in charge, specifically, are driven by religion and probably suppressing all dissent. It makes no sense at all to say “these few people, a subset of a subset, are theocratic”. So of course you were taken to be making a claim about The Scientific Establishment more generally. Anyway, it seems very untrue that there’s anything like a theocratic prohibition on multiverse theories; the nearest thing to a theocracy in fundamental physics at present is the priesthood of string theory, and I think on the whole string theorists are for multiverses.

  You describe my writing as “complaining,” and my discussion of risks as “complaints.” These words have an obvious negative connotation. Do you consider discussions about AGI risk to be full of “complaints,” or are they “valid concerns”? Why do you choose to use demeaning words to describe my concerns, but not the concerns of others? Also you say “...it seems very untrue that there’s anything like a theocratic prohibition on multiverse theories.” I don’t feel there’s a “theocratic prohibition” on multiverse theories, but I do feel there is a bias against, and some censorship of, multiverse theories and other philosophical outlooks on physics. I described this bias as being “theocratic” in the way physicists with this bias criticize and sometimes suppress the physicists who discuss “why” quantum phenomena might happen. I would encourage you to ask any multiverse theorist if they’ve ever encountered this bias. (And yes, string theory is one particular multiverse theory. I don’t think I said otherwise.)

  • The focus on “moderators of the physics subreddits” seems weird. Who cares about the moderators of the physics subreddits? (And why would Less Wrong be a suitable place to discuss their alleged deficiencies?)

  My original post had only one sentence about reddit. Granted, reddit is not exactly a bastion of intellectual thought. However, it can be used to get a real-time impression of sentiment among a certain portion of society. I referred to “moderators of the physics subreddits” as an example of the bias against philosophical physics in action. I thought Less Wrong would be an appropriate place to discuss the bias against philosophical physics. The moderators were just a real-time example. I have no more desire to focus on reddit than you do. I didn’t know I would have to elaborate on that one sentence repeatedly in these follow-up posts/​replies.

  • In fact, that focus strongly suggests a specific scenario: you have some theory of physics, you’ve been trying to promote it on Reddit, it hasn’t gone down well, and you’re annoyed about that. And, unfortunately for you, in the great majority of such cases the problem is not primarily with the Reddit moderators. Of course you might be the exception! But the expectation you’re setting up is not a good one.

  Again, the word “annoyed,” is like the word “complained.” And again, would you describe people worried about AGI risk as being “annoyed” by it? Or would you choose words like “concerned” and “worried”? And there was indeed a specific scenario or reddit that disturbed me recently, but it didn’t have to do with me “promoting” an idea. It was a post in which someone asked if Brian Greene’s perspectives were “BS” and there was pile-on of people criticizing Brian’s ideas. Obviously, he doesn’t need me to defend him. Still, the scenario represents the problem I’m talking about. The issue is not a personal grudge, but rather it’s a broader issue of a bias against larger, philosophical outlooks in physics.

  • Taking a look at e.g. /​r/​physics, there is discussion of philosophy, multiverses, etc., from time to time. So if the Evil Theocratic Moderators are suppressing such discussion, they aren’t doing it very carefully.

  I don’t think moderators are “evil.” Again, “evil” is an extreme and loaded word. Why not “biased”? Furthermore, your choice of capitalization in “Evil Theocratic Moderators” seems like an attempt to make me seem like I subscribe to something crazy and extreme. I imagine that it’s easier to paint me as extreme, and then criticize me for the alleged extremism, than it is to criticize any of my actual points. However, I would agree that this is just a bias I’m talking about, and not an all-out total censorship of all philosophical physics theories. There are indeed philosophical theories that make it through “from time to time,” as you noted.

  Also:

  • I don’t think shminux’s assumption that you could stand to learn some fundamental physics rather than handwaving about it was unreasonable, whether or not it was correct. Your original post doesn’t read like it’s written by someone with substantial expertise.

  As far as the advice to learn fundamental physics goes, I only mentioned my books on special and general relativity to convey that I do have an understanding of “some fundamental physics.” I would agree though that a person can always think more and more deeply about the basics. (For example, concepts like mass-energy and spacetime are almost infinitely interesting.) Also, I didn’t intend for my post to be like an academic physics paper. LW doesn’t seem like the right forum for that. I was only trying to discuss the idea of a supra-universal, philosophical approach to some physics phenomena, and the bias that exists against that approach.

  • Having self-published books about relativity is not strong evidence about whether a person has expertise in fundamental physics. (I actually suspect it may be evidence against.) Again, I am not claiming that you are a crank; I’m not in a position to judge. But the evidence we have available at present doesn’t give us much grounds for confidence that you’re not.

  This is getting tiresome, but as before, your choice of wording seems a little passive aggressive when you say “I am not claiming that you are a crank; I’m not in a position to judge. But the evidence we have available at present doesn’t give us much grounds for confidence that you’re not.” Words like “crank, annoyed, complained” seem slightly disrespectful. However, you’re right that, in my original post, I also was disrespectful when I said “I’ve noticed a universe-centric perspective among some scientists that seems to almost amount to a sort of theocracy. For example, there is a subset of the quantum physics community who refer to themselves as the “shut up and calculate” faction. They dissuade people from asking “why” certain phenomena occur.” I agree that “theocracy” is a loaded word, like “crank, annoyed, complained” and “evil.”

  As far as my books and evidence are concerned, DM me if you’d like a free PDF copy of one of the books, and I’ll send it to you. Then you can have more evidence to judge my knowledge of “fundamental physics” for yourself.

  [EDITED to add:] Oh, one other thing: your profile gives a URL for your web page, but it is the wrong URL. The real one has .org on the end, not .com.

  Nice catch.… thanks! I make that mistake all the time.

  Finally, I’m not going to engage further, so please try to refrain from saying anything else that misrepresents or exaggerates my positions, as I won’t respond. If people want to know what I actually said in the original post, the follow up post, or this reply, all they need to do is look. In any case, I appreciated the discussion.

  [EDITED:
  *I edited my original reply. Originally, I had written about why “you downvoted my post” when you (gjm) actually said you did not downvote my post. Sorry for the mistake. I was working on too many projects and replies at once, all while copying and pasting from a separate document. In order to fix the mistake, I have since changed “you downvoted my post” wording to “my post did not make a good impression on you” wording. I’m not familiar enough with LW to know if the platform will show the inline edits when I submit this edit. I hope it does, but if not, please know that I initially was incorrect on the downvoting versus “not a good impression” issue. I think the rest of my comments still stand though.]

• amelia 21 Feb 2023 0:25 UTC

  2 points

  1

  in reply to: Henrik Karlsson’s comment on: Child­hoods of ex­cep­tional people

  It’s so sad that other teachers weren’t on board with the advanced topics. Some adults can’t stand it when you teach kids about topics that they don’t understand themselves. I think it’s because doing so makes the adults feel less superior to kids. Just know that you were doing the right thing (if the right thing means helping kids to love learning, and to not be afraid of any topic). And what a gift for your daughter with a second language! She is so fortunate.

• amelia 19 Mar 2024 18:25 UTC

  1 point

  0

  in reply to: Ann’s comment on: Claude es­ti­mates 30-50% like­li­hood x-risk

  @Ann, this is really useful. Many thanks! The deeper we dig into this, the better. Even if people disagree with the numbers, I think having an LLM/​AI as the source of these numbers might get more people to pay attention to ASI risk. Thanks again!

• amelia 11 Mar 2024 19:59 UTC

  1 point

  0

  on: Is an­a­lyz­ing LLM be­hav­ior a valid means for as­sess­ing po­ten­tial con­scious­ness, as de­scribed by global workspace the­ory and higher or­der the­o­ries?

  I didn’t say this in my OP, but people who deny AI could ever become conscious remind me a little of the doctor and parents in this video from The Onion (“America’s finest news source”):

  https://​​www.theonion.com/​​brain-dead-teen-only-capable-of-rolling-eyes-and-texti-1819595151

  [20240311 added link]

• amelia 3 Mar 2024 1:27 UTC

  1 point

  0

  in reply to: amelia’s comment on: How is Chat-GPT4 Not Con­scious?

  And of course there is the gold standard paper for AI consciousness that @NicholasKees made reference to in their reply: Consciousness in Artificial Intelligence: Insights from the Science of Consciousness

  Based on the paper, and my best understanding, the type of consciousness that it seemed ChatGPT4 was displaying in the examples would be computational functionalism, in that it seemed to be choosing the correct algorithm to apply in the example problems.

  Thanks again to @NicholasKees. I probably should have known this paper was “out there” to begin with, but I stumbled into this consciousness question while “playing” with ChatGPT4, and did not come at it from a research perspective.

• amelia 3 Mar 2024 0:47 UTC

  1 point

  0

  on: How is Chat-GPT4 Not Con­scious?

  Here are some thoughts on “consciousness,” and how it might apply to ChatGPT4 based on the transcripts of the four sessions I provided in my OP and my initial reply to it:

  The obvious classic sources would be Nagel and Chalmers. However, the work most applicable to this discussion would probably be Seth, A. K., & Bayne, T. (2022). Theories of consciousness. Nature reviews. Neuroscience, 23(7), 439–452. https://​​doi.org/​​10.1038/​​s41583-022-00587-4 .

  I should have started with definitions in my original post, but I wasn’t expecting more than one or two people to actually read the post. In any case, using the theories discussed in Seth and Bayne, it seems like, based on the example provided in the OP and the four examples in my initial reply to it, ChatGPT4 might be considered conscious by higher-order theories, global workspace theories, and integrated information theory, as well as re-entry and predictive processing theories—if we consider the prompt input to be like sensory input for ChatGPT (analogous to auditory or visual input for most humans). I’m obviously not an expert on consciousness, so I apologize if I’m misunderstanding these theories.

  I’ve never been one hundred percent convinced that ChatGPT4 is conscious, as I noted in my replies to users. It’s just that, at the time I wrote my OP, I was having a a hard time comprehending how ChatGPT4 could perform as it did based solely on next-word probabilities. If, by contrast, it was actually learning and applying concepts, this seemed like a sign of consciousness to me.

  At this point, if I had to make an intuitive guess, I would come up with perhaps a .7 likelihood that ChatGPT4--during at least part of the time of the example sessions—would fit at least one of the theories of consciousness discussed in Seth and Bayne.

• amelia 1 Mar 2024 21:10 UTC

  1 point

  0

  in reply to: mouse_mouse’s comment on: How is Chat-GPT4 Not Con­scious?

  That’s an awesome idea about running a huggingface model on my home machine. I actually took some unusual measures to obtain my current GPU, so I really want to make full use of it. I can’t wait to try this.

  You also made an interesting point about the difficulty in seeing how a person’s interior experience arises out of just neurons/​anatomy. It’s fascinating to think about.

  The Othello article is absolutely mind-blowing! It obviously pushes my confidence in ChatGPT’s potential cognition up higher again, but the question of what type of cognition, and what type of potential consciousness ChatGPT might possess only grows deeper with discussion.

  I haven’t used OpenLLaMa, but the fact that it’s so bad at reasoning is indeed useful for seeing what an LLM looks like that probably didn’t form a general model—and highlighting how ChatGPT may have formed one.

  All of this discussion paints a really complex picture, but I’m enjoying the complexity, so many thanks!

• amelia 1 Mar 2024 1:27 UTC

  1 point

  0

  in reply to: amelia’s comment on: How is Chat-GPT4 Not Con­scious?

  After submitting the above reply to my OP, I went through some of the links people suggested, read those documents, and read @mouse_mouse ‘s reply also, and those resources changed my confidence levels since posting the above reply. I explain this more thoroughly in my reply to @mouse_mouse ’s reply below (assuming you’re sorting by the time the top of each reply thread was submitted).