tesseractcat's recent activity

  1. Comment on Learning and studying things without taking notes in ~talk

    tesseractcat
    Link
    As someone who takes notes: It's less about remembering things in the moment, and more about being able to recall the knowledge later (months/years down the line). Otherwise I'll just forget. This...

    As someone who takes notes: It's less about remembering things in the moment, and more about being able to recall the knowledge later (months/years down the line). Otherwise I'll just forget. This is also why I take digital notes despite studies that show that handwritten notes are better for memory.

    23 votes
  2. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    I'm very disheartened to hear that you believe that I've been arguing in bad faith. Since you stated you don't wish to continue the argument, I'll end this with what I believe is the fundamental...

    I'm very disheartened to hear that you believe that I've been arguing in bad faith.

    Since you stated you don't wish to continue the argument, I'll end this with what I believe is the fundamental disagreement I have: AI as a field cannot distill what AlphaGo does into a set of strategies that can then be applied to a traditional Go bot (except, maybe, by treating it as a black box). We can't look at the weights and biases to see how it models the Go board, or how it identifies patterns. I believe this represents a fundamental lack of understanding of these deep neural networks.

    It concerns me greatly when people say we do understand them, because it is like someone saying we do understand how the brain (the brain of any species) works. We may understand the lower level structures, and the very highest level, but we have a tenuous grip on the middle level, where the thinking happens.

    This is a hard thing to debate because the terminology is confusing. In some ways, yes, we do know exactly how deep neural networks work. We know lots about how to train them, exactly how to run them, what the GPUs are doing, the precise matrix multiplications needed. Understanding these digital neurons doesn't let us play Go though.

    5 votes
  3. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    My main goal with the above post was to counterbalance a tendency I see where people are quick to dismiss any possibility of LLMs having consciousness, not to argue for it. As for my personal...

    My main goal with the above post was to counterbalance a tendency I see where people are quick to dismiss any possibility of LLMs having consciousness, not to argue for it. As for my personal beliefs, I lean towards panpsychism, where all systems have some sort of qualia/awareness, even if alien to us. In general I dislike the term consciousness, and prefer qualia. I do think defining consciousness in terms of awareness is a bit circular.

    4 votes
  4. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    Sorry about that, I think it's a bad habit I got from the 'reply' feature on various chat apps.

    Sorry about that, I think it's a bad habit I got from the 'reply' feature on various chat apps.

    4 votes
  5. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    (edited )
    Link Parent
    First, I don't think agency is a necessary condition for intelligence, so I don't think this would really prove it one way or the other, but no intelligences are entirely self-bootstrapped. Human...

    When I say they have no agency, I mean it in the sense that they can't initiate anything, or want or prefer anything. I would look at the Belmont report as a good starting point on the ethics of agency.

    First, I don't think agency is a necessary condition for intelligence, so I don't think this would really prove it one way or the other, but no intelligences are entirely self-bootstrapped. Human intelligence is fed 'input data' through it's genetics, evolution, and upbringing.

    As far as how they work, we know tremendous amounts about the neural networks that compose them, how learning works, the algorithmic implicit differentiation, and how the learning works.

    These are all things other than how they actually work when running inference. We know quite a bit on how to water them, what humidity they need, and the correct scaffolding to grow optimally, but little on how they actually bloom.

    Similarly, techniques like targeted knowledge injection don't work by interfacing with the deep layers of the network, where the actual intelligence may or may not be, it works by interfacing with the input and output layers.

    And regarding philosophy, that gives us the framework to assess anything that might present as sentient.

    Has philosophy closed in on a consensus for the definition and process for determining sentience? If so I would love to learn more.

    5 votes
  6. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    I'm not arguing something does exist, I'm arguing it's impossible to know, and therefore impossible to make a definitive claim of it's nonexistence. You're right, proving a negative is impossible....

    I'm not arguing something does exist, I'm arguing it's impossible to know, and therefore impossible to make a definitive claim of it's nonexistence. You're right, proving a negative is impossible. Similarly, if we assume "that the thing/action/whatever does not exist until proven", this is true for all consciousness except for our own subjective experience.

    3 votes
  7. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    The person I responded to is the one making the definitive claim of lack of consciousness, so they should be the one to precisely define it.

    The person I responded to is the one making the definitive claim of lack of consciousness, so they should be the one to precisely define it.

    4 votes
  8. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    Source? It's my understanding we understand the 'architecture', but very little about how it actually accomplishes the goals we train it to do. They can do things with little to no direction, they...

    We know a great deal about how they work

    Source? It's my understanding we understand the 'architecture', but very little about how it actually accomplishes the goals we train it to do.

    They can't do things because they lack agency

    They can do things with little to no direction, they may not do them well but I think the level of quality is different from being entirely unable.

    Philosophy has centuries of work on the subject of consciousness

    But very little work dealing with massive neural networks and their consciousness or lack thereof.

    And they don't learn from there output being fed back in

    They do learn (to some extent) within the context window, continuous learning is definitely an area of research though.

    6 votes
  9. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    We don't know how they work. We know how we trained them. It's more accurate to say an LLM is grown, rather than made. Maybe on the inside they are simple probabilistic engines, or maybe they have...

    These transformer networks are simple probabilistic engines working on tokenized text.

    We don't know how they work. We know how we trained them. It's more accurate to say an LLM is grown, rather than made. Maybe on the inside they are simple probabilistic engines, or maybe they have a more complex world model, or maybe they do any number of other things.

    They can't acquire knowledge because they can't do anything

    They can do things.

    they have no sense of self, or consciousness, sapience,

    These are all subjective things that are impossible to measure.

    sensory loop

    LLMs do have a rudimentary sensory loop in the sense that their output gets fed back into them.

    6 votes
  10. Comment on It’s OK to call it Artificial Intelligence in ~tech

    tesseractcat
    Link Parent
    Just want to say that we don't really know enough about consciousness to say this for certain. The only proof we have that anything is conscious is that we experience it subjectively. The only way...

    LLMs are not sentient, and they are not conscious.

    Just want to say that we don't really know enough about consciousness to say this for certain. The only proof we have that anything is conscious is that we experience it subjectively. The only way we can predict consciousness of other things is then by their similarity to us. LLMs are possibly the most similar artificial creations to humans (although that's not saying much), so therefore have a higher (but still low) chance of being conscious.

    3 votes
  11. Comment on What are your favorite aesthetics? in ~misc

    tesseractcat
    Link Parent
    Solarpunk is often drawn in an anime style, and Miyazaki is the director of many of the most popular anime movies. Additionally, Miyazaki's movies often have magic/slightly more advanced...

    Solarpunk is often drawn in an anime style, and Miyazaki is the director of many of the most popular anime movies. Additionally, Miyazaki's movies often have magic/slightly more advanced technology (as seen in Howl's Moving Castle and Nausicaa, for example), which is solarpunk-adjacent.

    11 votes
  12. Comment on [SOLVED] Recovering data in a very old, possibly corrupted tar archive? in ~comp

    tesseractcat
    Link
    Depending on how much time you want to dedicate to the task one strategy would be to open up the file in a hex editor and compare it to the GZIP specification to see how exactly it's corrupted....

    Depending on how much time you want to dedicate to the task one strategy would be to open up the file in a hex editor and compare it to the GZIP specification to see how exactly it's corrupted. Depending on the type of corruption it might be possible to fix it, or to extract some of the compressed data and decompress it separately.

    5 votes
  13. Comment on The Boys | Season 4 official teaser trailer in ~tv

    tesseractcat
    Link Parent
    Addressing a few points here. The show isn't really adapting the comics, it's more loosely inspired. The other heroes don't know that he can be beaten, and it's a very risky proposition to try...

    Addressing a few points here.

    The comic ran 72 issues, and I'm not sure where the show is in relation at this point.

    The show isn't really adapting the comics, it's more loosely inspired.

    The other on-screen heroes are apparently just terrified of him; too scared to band together and plot his downfall to clear the way for their own rise to power.

    The other heroes don't know that he can be beaten, and it's a very risky proposition to try because if you fail that's it.

    Because if Homelander and the others actually were as narcissistic and megalomaniacal as they're written (and act until they hit the story bumpers), Homelander would be ruling rather than just sulking that his follower counts aren't higher. Homelander, as written, would have pretty much vaporized at least the entire American government until he got to the people who slavishly fawned over his every whim, and then sat on his throne of skulls basking in the glory of being the God in Charge.

    I don't think this would be the case. Homelander is a complex character, and one of his motivations is to be seen as "good" and a "hero", rather than a mass-murderer.

    I do think Homelander's arc is ready to be resolved though... Season 3 felt like it was spinning it's wheels.

    17 votes
  14. Comment on Portal: Revolution | Trailer in ~games

    tesseractcat
    Link Parent
    If you're looking for new mechanics, you might enjoy Portal: Reloaded (https://store.steampowered.com/app/1255980/Portal_Reloaded/), which is a mod that adds a whole new campaign with a new type...

    If you're looking for new mechanics, you might enjoy Portal: Reloaded (https://store.steampowered.com/app/1255980/Portal_Reloaded/), which is a mod that adds a whole new campaign with a new type of portal (time travel portals).

    13 votes
  15. Comment on Portal: Revolution | Trailer in ~games

    tesseractcat
    Link Parent
    This is a free to play, fan made mod, not an official sequel.

    This is a free to play, fan made mod, not an official sequel.

    19 votes
  16. Comment on What cheese makes the best Mac & Cheese? in ~food

    tesseractcat
    Link Parent
    Would you mind elaborating? Maybe link some stuff you prefer?

    Would you mind elaborating? Maybe link some stuff you prefer?

    7 votes
  17. Comment on How did deepfake images of me end up on a porn site? in ~life

    tesseractcat
    Link Parent
    I think it's questionable how much you can maintain anonymity and still regulate individual internet usage... This is why you see so much backlash to even small regulation (it's slippery slopey)....

    I think it's questionable how much you can maintain anonymity and still regulate individual internet usage... This is why you see so much backlash to even small regulation (it's slippery slopey). All part of the larger war on general purpose computing, which is great if you're a victim of deepfakes but not so great if you're posting anything subversive.

    12 votes
  18. Comment on GPT-4 understands in ~tech

    tesseractcat
    Link Parent
    LLMs have a pretty good internal map of word length, but it's not perfect. You can actually get pretty close if you ask it to tell you the length of all the words in a sentence, and then ask it to...

    LLMs have a pretty good internal map of word length, but it's not perfect. You can actually get pretty close if you ask it to tell you the length of all the words in a sentence, and then ask it to add them all together. It will make a few mistakes though, like including punctuation in words sometimes (which can occur because it sees tokens not words), or just getting the lengths wrong. It's actually pretty impressive how well it can do considering it doesn't really see the characters, instead its learned the lengths somewhere in it's training data.

    3 votes
  19. Comment on GPT-4 understands in ~tech

    tesseractcat
    Link Parent
    LLMs don't actually see words/characters, they see 'tokens'. Whether or not LLMs can understand things aside, this wouldn't be a good measure of it. It would be like asking a Chinese speaker how...

    LLMs don't actually see words/characters, they see 'tokens'. Whether or not LLMs can understand things aside, this wouldn't be a good measure of it. It would be like asking a Chinese speaker how many characters long a sentence was in English (given a sentence in Chinese).

    8 votes
  20. Comment on Database containing nearly 200,000 pirated books being used to train AI - authors were not informed in ~tech

    tesseractcat
    Link Parent
    I agree, it's mostly predictable that it didn't work out so well. I still think there's value in trying different incentive models (even if they don't work out), as private trackers can be quite...

    I agree, it's mostly predictable that it didn't work out so well. I still think there's value in trying different incentive models (even if they don't work out), as private trackers can be quite restrictive.

    3 votes