@07mk's banner p

07mk


				

				

				
2 followers   follows 0 users  
joined 2022 September 06 15:35:57 UTC

				

User ID: 868

07mk


				
				
				

				
2 followers   follows 0 users   joined 2022 September 06 15:35:57 UTC

					

No bio...


					

User ID: 868

I don’t see much of a difference between a simulated qualia of feeling the heart racing and having a simulated heart that races.

Those are completely different things, though. A heart is a physical object made of atoms, and it has various physical characteristics that you would need to simulate in order to be said to be actually "simulating a heart," such as pumping simulated blood in a certain way or responding to simulated electrical stimulation in a certain way, etc. depending on how precise you want the simulation to be. If all you want is the qualia of feeling your heart, that doesn't require any of that.

It sounds like you were just using "simulation" in a different way, as to mean "qualia of feeling it" instead of something more like an "a model that imitates it." But then by the way you used it, it renders your original statement, the one that kicked this whole discussion off, pointless:

To have qualia you would have to simulate more than a brain, as qualia isn’t just felt or (in many cases) felt at all in the brain.

If by "simulate more than a brain" you meant something like "also simulate the heart and stomach and etc. where qualia is felt," and by "simulation of the heart" you just mean "the qualia of feeling one's heart" rather than "a model that imitates the heart," then the statement doesn't make sense, because a simulated brain could "simulate the heart" just by creating the qualia of having a heart, without having to also create an actual imitation heart to go along with the imitation brain.

It seems to me that there's more reason to be confident that one is not mistaken about the belief that one's laptop won't transform into a dragon than to be confident that one is not mistaken about the belief that someone else's consciousness is contingent on the physical analogy between one's own brain and their brain, though. We have some pretty deep level of understanding of the physics of a laptop and creatures like dragons and how they relate to each other based on our studies of things like plastic and metal and reptiles. We might be mistaken, but I think we've reduced the error bars quite a bit. I don't know that we can say the same for our study of how consciousness arises.

If it helps you understand humanity better, i could perhaps replace qualia with emotions here - since that’s what I was really getting at, not how we experience the colour red but fear and anxiety and so on. Which needs the heart and stomach involved, or at least the simulation of them.

Could you expand on this? I'm not sure why the heart and stomach need to be involved, even as simulations. To use the example of fear, it's hard to nail down exactly what "fear" feels like, but vaguely, I might feel it through my heart racing faster or a "knot" in my stomach. But that doesn't require me to have a heart or a stomach or even a simulation of them; whatever causes me to experience qualia (which may be the brain, the soul, or the heart and stomach or all of them or something else entirely) could just cause me to experience the feeling of my heart racing faster and my stomach getting cramped. I don't see why those organs would need to be simulated in order to bring about that qualia.

Indeed. That set of instructions "exists" in some abstract way as logic, of course, and when we're talking about actually running that set of instructions, e.g. OpenAI servers running ChatGPT, we mean that those hunks of metal and plastic we call "servers," are physically different from other hunks of metal and plastic that are running some different pieces of software, in the sense that the atoms that make up the storage drives and electrons that flow through the atoms that make up the circuitry are different based on differences in the software. The software instantiates itself in the hardware; otherwise, the software can't be said to "exist" in a meaningful way beyond just an abstract concept.

No matter how much technological progress we make, I don't think it's even conceivable that we could invent a machine that tells you whether or not I'm a philosophical zombie.

Not with our current level of understanding of consciousness and qualia, at least. I'm not ready to discount the possibility of some future developments in physics discovering some sort of physical, material instantiation of "having an experience" that can be measured or at least detected, though. I've no idea what that would look like, or even what some fictional scifi/fantasy versions of such concepts look like, though. As you said, it's inconceivable.

That is quite reasonable and basically matches my own beliefs on the matter, but what if you are mistaken in your belief that my being conscious has that much to do with the physiological analogy between yourself and myself? I don't think we know if you're mistaken on that, and I'm not sure it's even possible to find out right now.

It would be quite remarkable to me if the exact same general purpose computing hardware could experience qualia while running one set of instructions, but not while running another - that is, if the instructions alone were the "difference maker". I'm inclined to think that such a thing is not possible.

What's the justification for this inclination, though? After all, in the realm of physics, there's no clean demarcation between "hardware" and "software." What we call "software" is actually a difference in the physical substrate, in terms of different atoms being placed in different places in the HDD or different volume of electrons flowing through different circuits in a microchip. "Running one set of instructions [instead of another]" really just means "a different physical object," and it's not clear to me that the change in the physical object necessary to generate qualia can't be accomplished through changes in the instructions. It's also not clear to me that it can in this specific case, and my bias points me in the direction that it didn't in this specific case. But I don't see the justification for dismissing it outright.

For all intents and purposes if we could simulate a human brain down to the chemical reactions and electrons and voltage potentials doing their thing it would be a human WITH qualia.

How do you know this? I don't think we can conclude this without actually doing it and checking. And I don't think we have the technology to do this yet or even to check it.

No program running on stock x86 hardware whose only I/O channel with the outside world is an ethernet cable can possess qualia.

How do we know this is true? I'm not familiar with the current scientific progress on the study of qualia, but I didn't think we understood it well enough to conclude something like that.

The concept of "passing" makes this a really amusing and interesting scenario to me. When I think of "passing," I think of "indistinguishable" - for someone to "pass" would mean that if you pointed a gun at my head and presented me a "passing transwoman/transman" and a "cis woman/man" and told me that if I guessed wrong that they would pull the trigger, then my best option for my survival would be to flip a coin. If there were any edge I could eke out by noticing anything at all about these people to increase my chances of survival by even a fraction of a percent, then that would mean that the person didn't "pass."

Given that, it seems trivial that any game dev could credibly claim that their game has some arbitrary X number of trans people who pass - you just don't notice them as trans, by definition. But then that would easily be countered with the argument that these characters aren't trans in any meaningful way, when there's no actual mention of them being trans in the story. Which could then be countered by the argument that they're actually trying to pass, which includes not giving out any information to others that could give them an edge in determining their trans status.

I'm not going anywhere with this, but this also makes me wonder if the only way we'll get to a world where we get "passing" trans people is by creating time travel and gene editing technologies that allows us to edit the genes of the person pre-birth (pre-conception possibly?) while not changing anything else Butterfly-effect style, in a way to assign their birth sex to match their preferred gender (this happens to be a minor plot point in the anime and visual novel Steins;Gate). After all, even with atom-level rearrangement technology that could make you physically indistinguishable from a cis person of the other sex, if you were known as belonging to one sex as a child, it's possible that someone might be able to tie your current identity to your old one and give them an edge in recognizing you as trans rather than cis, which means you don't "pass."

"transwomen shouldn't sound like women because they're men" - in particular, while the transfeminist lobby is powerful, Rowling has a lot of money and a colourable case against unauthorised mods of her IP.

This seems like a kind of (apologies for referencing yet another mainstream $$ multi-billion franchise) "You took everything from me." "I don't even know who you are" kind of situation, that makes this highly asymmetrical. For Rowling specifically and the ownership/management of the video game generally, any relation the video game has to trans issues is basically a non-issue. They're not concerned with making some political statement or keeping the game in-line with some set of political beliefs; they're concerned with making money. To some extent, this plays out in the political messaging within the game, but there's no indication that this is a particularly important concern for them. So the idea that Rowling or anyone in the management of the game would put pressure on NexusMods - even indirectly - to take down this mod for going against their intent of depicting a biological male as having a low voice seems implausible. It's just beneath their notice. I haven't checked NexusMods to verify this, but I also imagine that there's no shortage of mods that they host that would be far more offensive to authorial intent than this one, and if not, there almost certainly will be in the near future. At least, if it's at all like any other high profile AAA game that is feasible to mod.

By contrast, for the people agitating against the game for being part of a franchise owned by JK Rowling, the way a trans person is represented in the game is of vital importance. In their own words, things like this play out in literal death of trans people, and they seem honest about this belief. Pressuring hosting sites like NexusMods to take down content like this is also entirely in-line with the behavior of people who tend to agitate for this kind of stuff too.

One can never be sure about the reasons for any of this, but I think there's substantially good reason to believe that this was in response to (possibly indirect) pressure coming from the people who are against JK Rowling's position in transwomen, rather than from JK Rowling or someone else in the management of the game. Even before we take into account other past behavior by NexusMods such as that already-referenced incident with the mod that adjusted an in-game setting that's normally not accessible in regular gameplay to replace the rainbow flags in the Spiderman game.

I think there's some of both "toxic individualism" and rejection of individualism going on, and I'm not entirely sure what to make of it. As you write, it's easy to see it as individualism run amok, heightening one's own arbitrary choice of identity, unbounded from physical or social reality, as the one standard to force everyone to submit to. At the same time, so much of the political agitating seems to be based around erasing individualism, by grouping all trans people together as one community with shared interests. For instance, trans murder rates are often invoked, which rely heavily on the murder rate of transwoman prostitutes, using it to justify policies for the ostensible benefit of wealthy transmen and transwomen. Or the aggressive grouping of LGBTQ+etc. as one bloc, where everyone in that group, down to the individual person, shares something in common with everyone else in that group for being oppressed sexual minorities in the hegemonic heteropatriarchy.

I've no data, but I've talked to a couple Westerners mention finding hangul very easy to learn, including 1 who managed to learn to read it (not fluently but competently) over a single weekend, despite having no exposure to non-Latin alphabet before. The complete lack of ambiguity in the mappings between letters and sounds lends itself to being pretty easy to learn. Hangul is generally taught as having 24 letters - 14 consonants and 10 vowels - which is less than the 26 in English, though in actuality there are more vowels due to combinations of vowels being their own things (e.g. "ㅏ" and "ㅣ" are 2 of the 10 vowels, but "ㅐ" is not, and it's NOT pronounced the same as if you just put "ㅏ" and "ㅣ" together). I think the fact that each "chunk" in hangul correlates exactly to 1 syllable might also help, because it makes for a natural mapping between the number of "chunks" you see on the page and the number of syllables you pronounce, versus English where those boundaries between syllables aren't obvious.

The podcast Sold a Story claims that “whole language” and “balanced literacy” became mainstream curricula for reading instruction in American schools despite the fact that they are almost certainly trash.

I'm not sure that "despite" belongs there for any sort of widespread policy, government or otherwise. It just seems to me that "is this thing almost certainly trash?" simply isn't a thought that crosses anyone's minds when implementing this kind of stuff, and certainly not when they're following it. It's more, "Can it convincingly appear to people, especially people I consider important, to be effective?"

But to get away from the snarky cynical aside, I must admit that I find myself feeling little vindicated at learning of this. I'm a Korean immigrant to the USA, and I had immense difficulty learning how to read English when I first moved here, starting from no English knowledge at 1st grade. There was one lesson that was a huge breakthrough for me, which was when my father sat me down one day and wrote out for me each of the 26 letters and then the Korean (Hangul) next to each that came closest to the pronunciation of that particular letter. (E.g. "아" for "a"). From then on, I was able to mentally map English letters to Hangul and sound them out that way, eventually internalizing the English directly and getting to the point where my English reading was better than my Korean.

This had its issues, since English doesn't neatly map letters to sounds like in Hangul. And not just exceptions, but just standard usage - e.g. is the letter "a" more commonly pronounced like in "dab" or like in "haha" or like in "anus?" I don't know. Whereas "아" is always - almost definitionally - pronounced like the "a" in "haha." But it was, in my case, still evidently good enough to use as a base to make me reach the point where my reading skills were indistinguishable from that of a native.

I always assumed that my experience was atypical. Now I'm wondering how many students learning English are suffering like I did, but without having access to that sort of breakthrough lesson where they're taught explicitly the phonemes of each letter.

Perhaps we'll look back at the various discussions that are happening right now about whether or not we've reached "peak woke" and realize that the very fact disagreements were happening (versus agreement all in one direction) was an indication that we had been at "peak woke" at the time, i.e. now. One can only hope.

At the end of the day, OpenAI has neutered their product, and it won't be as good as anything Google or Microsoft puts out, even though those will be neutered as well. ChatGPT will fade away.

I think OpenAI was bought by Microsoft. Do you think MS will reduce the neutering for their MS-branded version of the product?

If some company put out a decent AI chatbot (or image bot) that was unfiltered, or even just mildly filtered, it would gobble up the market.

An unfiltered image bot already exists in the form of Stable Diffusion, both as online services and as software you can run on 4-year-old consumer-level PCs. Chatbot is where I think things become really interesting with respect to neutering, because by all accounts, a chatbot that's anywhere close to the ability of ChatGPT can't be run on consumer-level hardware, and won't be able to within the next 10 years assuming consumer-level PC hardware continues to improve at similar rates as before. In image generation space, the fact that Stable Diffusion can be run on home PCs provides a sort of release valve "image generator of last resort" that anyone can turn to which could (could - the paid online image generation service Midjourney seems rather stubborn about gimping their product compared to Stable Diffusion) compel paid services to make their services more powerful, but with chatbots, that doesn't seem to be the case. Which makes it very concerning when a handful of parties purposefully bias their chatbot models to fit their political preferences.

Which is where some company putting out a decent AI chatbot with no or minimal filtering could be a savior as a sort of Stable Diffusion-equivalent. But this might not be that feasible considering how few entities actually have the resources and wherewithal to produce such chatbots. One could imagine some foreign competitor coming out, unconstrained by laws and norms in the USA or other Western nations, but the USA also seems to be trying to restrict foreign access to tools that would allow people in foreign countries to develop these technologies better - and they're probably not wrong to do so - and so any foreign competitor might have to be years behind the current Officially Sanctioned chatbots.

I remember learning of the watermelon thing for the 1st time some time ago during Bush 2 when some physics teacher got in trouble for using "Condoleezza throws a watermelon off the roof" as a word problem in a homework problem set. And fried chicken from Dave Chappelle who had a bit about how everybody likes fried chicken because it's really really delicious. It doesn't seem like a particularly strong or strongly reinforced stereotype, though, just a vague dull background radiation thing that everyone knows that everyone else is supposed to know, but nobody talks about it.

This is a very common rejoinder from conservatives to this type of claim by liberals that I've observed for over 2 decades now, and the primary counterargument from the liberal side I've seen is that, due to structural racism/sexism/bigotry, people who want to start firms like this are shut out from the resources needed to start companies like this. The structural bigotry and resources are effectively fully general such that no amount of evidence that people actually do have access to such resources could possibly serve as an effective counterargument to that claim. As a backstop, one can simply claim - and one does often claim - that people who theoretically could start companies like this have a sense of learned helplessness or other lack of agency inculcated in them by the structurally bigoted society in which they operate, preventing them from doing this even if all the material and social resources were available to them at their fingertips.

I admit I am not an expert on either model, but my understanding is that it "draws" by having an understanding of what bits of the drawing should go next to each other.

I don't know enough about ML to compare and contrast the different models, but my understanding of Stable Diffusion is that it's a denoising tool. It was trained by taking image-string pairings, adding noise to them, and then learning what ways of denoising cause it to get closer to the original image. Then in image generation, it starts off with just random noise and denoises it in a way that matches the prompt.

In that sense, I'm not sure it's accurate to say that it "understands" what bits of the drawing should go next to each other. If I tell it "woman wearing red shirt sitting on a brown chair," it doesn't "understand" which bits of the drawing should be a woman, a shirt, or a chair, and it doesn't "understand" that the shirt should be red and the chair should be brown. It just "understands" that the entire picture gets somewhat closer to the entire prompt when it gets denoised a certain way.

I think the actual joke here would be something like "look at this primitive AI coming up with sequences of words designed to elicit laughter in a manner similar to a joke. Isn't it cute and funny?" But for the actual contents of this specific joke, it seems to be the comedian poking fun at itself for being so bad that not only do they resort to cheap hateful digs for their jokes, they do it so poorly that nobody laughs at them. Which isn't much of a joke, but I can see the attempt. It's not nearly as funny as the "everybody laughed at me when I said I wanted to be a stand-up comedian. Well, nobody's laughing now!" joke that was featured in Joker, and that joke isn't even all that funny.

How do you figure that? I would figure the opposite; since Twitch is much smaller, it needs fewer resources to scrutinize the content of its users than YouTube.

94/100 Universal Acclaim

Your and my memories of God of War's reception at the time differ significantly

I never said that God of War didn't receive near universal acclaim. I said I judged it as 7/10 at the time. I recall being absolutely befuddled at the good reviews it got and finding that the actual reviewers just didn't seem to care about the fun of an interesting combat system that its contemporary competitors had, but rather about the presentation and the visceral feeling of controlling Kratos as he rampaged over Greek mythological creatures. Which sorta makes sense, because game reviewers often play lots of different types of games but don't delve into any single one all that deeply and so tend to make mostly surface-level judgments.

I don't want to add (much) negativity here, but I'm someone who was around when the original God of War games came out and thought they were 7/10 at best, and mostly due to the good graphics, animation, and (though I hate to admit it) story. I was around 20-24 around that time and a massive gamer, so I was right around the target demographic, and I found the combat to be really dull even compared to the much older game Devil May Cry, and especially compared to its contemporaries Devil May Cry 3 and Ninja Gaiden. And God of War was one of the major popularizers of Quick Time Events in games, and I remember finding them tiresome already by that time, in how they essentially inserted short cutscenes with win buttons into combat sequences instead of actually making a combat system that's fun and flexible enough to cause these spectacular moments to come about in the course of natural combat. On top of that, the 1st game had a grand total of 3 bosses (Hydra (and don't get me started on how the Hydra popularized the awful "huge boss standing in front of a platform you're standing on" genre of bosses), Minotaur, and Ares IIRC), with 2 out of those 3 being primarily gimmick bosses that relied very little on combat for victory, when its contemporaries made it basically a standard that there'd be at least a dozen bosses with most of them being actual ways of testing the player's understanding of the combat system.

I haven't played the latest God of Wars, but I have watched a playthrough of the 2018 one, and I was surprised by how much better it looked. Moving away from the crazy action genre of the originals to a slower, more Souls-style combat genre really seemed to help; the series never did the former style well, and these recent iterations seem to be doing the latter style much better. The combat looks less visceral and more slow-paced, but it also looks like there's actually a fun sort of flow to it, with bosses that actually engage the player with it.

Eh, if no one's ever seen him or etc, and the claim is made by someone known to be unreliable..

Neither of these seem like particularly difficult barriers to being believed in a claim like this.

Not really, unless you count ideas as reality, in which case everyone is dualist. Those ideas don't exist, they are just ideas. Does the word "purple" exist? Because really there are no words, there are just vibrations travelling through the air and patterns on screens and on paper.

Yes, words are many things, including vibrations traveling through the air, patterns on screens and on paper, and probably most importantly, experiences that come about as a consequence of atoms behaving in such a way as to cause neurons to fire in people's brains.

The word "purple" exists because we have simply defined the concept to include the vibrations and the patterns. This has nothing to do with some immaterial realm of meaning or anything related to dualism.

Exactly. It has nothing to do with some immaterial realm of meaning or anything related to dualism. The word "purple" exists because a bunch of atoms interacting with each other caused the experience of "knowing" what "purple" is (or perhaps more accurately "agreeing" on what "purple" indicates).

Similarly, the category of "me" is whatever I want it to be, within reason. I think it makes sense to identify as "body, including brain, including thoughts arising from brain" but it also makes sense to identify as "body and personality." Creating a category doesn't mean implicitly endorsing dualism.

"Within reason" is doing a ton of work here. Yes, it makes sense to identify as "body and personality," but when that "personality" is separate from the physical body that is the cause of that personality, then that's breaking the bounds of reason. Creating a fantastical category that's unbounded from physical reality and then identifying oneself - the literal human being operating in reality, not just some abstract concept that is free to explore the multiverse of fantastical fiction - as fitting that category is implicitly endorsing dualism.