Using the reasoning given by unenlightened, on day 100 every blue-eyed person would leave knowing that they have blue eyes, every brown-eyed person would leave knowing that they have brown eyes, and the Guru would stay knowing that they have neither blue nor brown eyes. — Michael

What's the reasoning for brown eyed people? Unenlightened gave reasoning for blue-eyed people

oh wow! you're a true hero

Now @unenlightened @LuckyR, here's the more tricky part - what new information did the Guru give them that they didn't already have?

☝

You two guys are good, I actually couldn't figure this one out myself.

Farmers get the benefit of socialism even if they vote against socialism, so they reap the rewards of socialism for themselves and deny it to everyone else.

(And I'm using the word socialism here loosely as, benefitting from money gained from taxes - US farmers are heavily government-subsidized)

So it's a win-win for them - all the benefit for themselves and no one else. Very cool farmers.

it means paste your reply into this website and click encode

https://www.base64encode.org/

And send me the result that comes out below

There's a DECODE button at the top of the site to do the reverse process

QW55d2F5LCB5ZXMgdGhhdCdzIHRoZSBhbnN3ZXIhIFlvdSBmaWd1cmVkIHRoYXQgb3V0IHdpdGhvdXQgbG9va2luZyBpdCB1cD8gSW1wcmVzc2l2ZQ==

maybe base 64 encode your post and I can base 64 decode it

[spoiler] testing testing [/spoiler]

that colour thing didn't work. I was wondering if this forum has SPOILER technology

nah the solution doesn't need to involve any work arounds like guessing and failing. There's a clean solution.

I included this text from the author at the bottom:

There are no mirrors or reflecting surfaces, nothing dumb. It is not a trick question, and the answer is logical. It doesn't depend on tricky wording or anyone lying or guessing, and it doesn't involve people doing something silly like creating a sign language or doing genetics. The Guru is not making eye contact with anyone in particular; she's simply saying "I count at least one blue-eyed person on this island who isn't me."

All the people except the guru have a pretty good idea what color their eyes are. After all they can count. — T Clark

They can count, but... so what? What's the logic? From the point of view of any person showing up to the boat, how has he logically deduced the colour of his own eyes?

every detail given is part of the solution. The fact that they all have the opportunity to leave exactly once a day is relevant to the final solution - the answer to the solution will involve specifying what night they leave eg first night, second night etc.

none of these match the canonical answer but I would love to see your justifications anyway

argument against it. But I still see a difference that feels important. DNA, I suspect all natural information, has an objective goal. — Patterner

I would wager that you think it's special and different because it's part of you, and you think you're special and different

DNA is just chemicals. And no, it's not "Unless something prevents it, something specific will come of it" - that's not true of DNA. DNA requires a very specific environment, and very specific resources, to do what you are calling its "goal". It doesn't need to be actively prevented to fail - it needs to be actively enabled to succeed.

Take a bit of DNA outside of it's special little environment and watch it do nothing.

it seems fairly obvious to me it's processing information. No?

You accept that DNA is processing information - DNA is chemical dominos as much as anything computers do, including LLMs. If DNA is information processing dominos, LLMs can be too

interesting question for you:

Physicalism aside, if consciousness is fundamental, is there something it's like to be an LLM?

If physicalism is true, then an LLM that knows all the physical facts about a book knows everything there is to know about the book. — RogueAI

And this still doesn't follow


There's no "if" when it comes to LLMs. Physicalism IS true for LLMs. They don't have souls. Everything they do and "know" and "understand" is all physical.

This is tricky because we're still not at sure whether an LLM will ever be able to understand anything, — RogueAI

But you wouldn't be sure no matter what, period. There's no possible world where you would even admit the slightest possibility that it's understanding.

You can literally, right now, give it a text it's never seen before and ask it for a summary and it will do a damn good job. Even in the face of this you won't give any ground, meaning you're not the kind of person to give ground on this period, no matter what, in any particle world

current LLMs convert inputs to outputs, just without muscles or physical sensation. Current LLMs make sense if signals - that chess gpt link demonstrates that well, we can see that they convert a series of chess notation into a chess board state. They do it without physical sensation, and without motor actions, sure, but they still do it.

I don't think they're conscious yet.

That's not the point of this conversation anyway. Rogue basically said, if everything is physical, then you should be able to understand the meaning of a book by just having physical access to it, even if you don't know the language it was written in. Which seems... absurd to me, to be honest. And we have what we could reasonably consider something not too far off from "physical understanding machines" in these LLMs - they display all possible outward signs of understanding. They're perfectly physical, and yet if you gave them access to a text written in a language they're not familiar with, they won't understand it. I consider that to be essentially tangible falsification of rogue's idea.

Do you think LLMs understand text? I don't think they have the slightest understanding that the marks on paper, or the binary code that the marks on paper are converted to, mean other things. I don't think they understand what meaning is, even when they are programmed to say they are. I think the binary code reacts in different ways to different binary code that is input, entirely determined by how they are programmed. I think it's very complex dominos. — Patterner

I think it's the only tangible comparison we can make at the present moment. Whether they "truly understand" or not is... kinda inaccessible to us. They pass the turing test, they give us all the signs we would expect of understanding, and so... as far as I'm concerned, it's the most valid existent comparison point to human understanding.

They do this:
https://www.lesswrong.com/posts/yzGDwpRBx6TEcdeA5/a-chess-gpt-linear-emergent-world-representation
They internally represent what "they think" the "world" looks like. If that's not some attempt at "understanding" I don't know what is.

And so we can say, "Maybe we don't know if human minds are physical or not, but we know for sure LLMs are physical, and they display all the signs of understanding, including internal representations of what they 'think' the state of the world is, so... you can't just blanket say 'if human minds are physical, then they would understand every language in the world if it were written down'" all the stuff Rogue was saying. What he's saying is pure speculation (probably worse than speculation, tbh, it sounds like gibberish to me), LLMs are the closest thing we have to non-speculation about the topic of understanding. Obviously it's speculative to some degree, but it's decidedly less speculative.

I don't think the Chinese room argument is very good, to be honest. I think it misses the point entirely.

I'll check out the second one.

So anyway, the claim now from you is, if physicalism is true then knowing everything about the physical arrangement of the book should allow you to understand the meaning of the book, even if you don't understand the language it was written in.

I just don't think that follows.

I mean, let's take LLMs as an example. They're a good example because they're explicitly physical. They are implemented 100% in the physical world - the computer scientists who invented them didn't learn how to imbue them with souls or anything, they work on the same physical principles as any normal computer.

Now if you give one of these LLMs a bunch of text in a language they're trained on, they can summarise it for you pretty well.

And if you give them a bunch of text on a language they haven't been trained on, they can't.

So we have a fully physical system which can, loosely speaking, "understand" some stuff and not "understand" other stuff, despite having the same access to the visual characters of each text. So... no I don't think it holds that, if physicalism is true, a person should be able to understand text he hasn't been trained to understand.

Obviously LLMs aren't the same as human beings and a summary from the LLM isn't the same as human understanding. BUT the ability to summarise and paraphrase a text is a human test for understanding, so I think the comparison is honestly robust enough.

ah I didn't realise you were including language issues in that. The way you phrased the question made it sound like you thought there was a difference between knowing everything about the book Vs physically having it, whereas now the problem is really knowing the language of the book, Vs not knowing the language.

ETA2: If physicalism is right, then a book is just ink on paper; patterns of squiggles. So a person with total physical knowledge of a book (ink chemistry, paper fibers, locations of atoms, etc.) should, in theory, know everything about the book. — RogueAI

This one is rather trivial. Of course someone with that knowledge could in principle learn anything about the book someone who physically had the book could. They'd have to do more work than someone who just had the book in front of them, but... so?

I think he wants debates within panpsychism. Which is valid. "If we start with the assumption that pansychism is true, where does that lead us?" I think that's fair. I'm not a panpsychist myself but I think that kind of approach is worth having. It's an exploration of an idea.

are you the guy who listened to the Annika Harris audio thing with me?

there's a big difference between not having convinced anyone at all, and not having convinced you in particular

maybe there are lawless universes. And if there are, interesting stuff doesn't happen there and there aren't conscious beings there to wonder why it's so lawless.

I don't think there are lawless universes though, I think a universe is fundamentally defined by its laws.

but they would happen without a timeline — ArtM

that seems like an assumption, not a fact

Imagine waking up tomorrow, realizing that thirty years of your life vanished, not forgotten, but as if they never existed at all. You jumped from infancy to adulthood in the blink of an eye, with no memories in between. This scenario sounds impossible, yet it’s exactly what occurs in situations like comas, alcohol-induced blackouts, or even during periods of deep, dreamless sleep. Here’s the profound question that emerges: if time is genuinely a fundamental dimension of our universe, why does it cease to exist the moment consciousness fades away? — ArtM

Disappointing opening paragraph. Why does it "cease to exist"? Just because some person isn't perceiving it, that means it "ceases to exist"? Hmmm... I'm pretty sure if a person is in a coma for 10 years, their body shows all the signs of time continuing to exist...

Maybe OP lacks object permanence, or is some kind of solipsist. Otherwise I can't really make sense of this.

For example ''Superman can fly in the fictional realm of DC''. Is true if stated as such and thus is knowledge. It doesn't require a belief to be true. It just is. — Jack2848

Nah, it has to involve belief. It's not "knowledge" unless someone knows it, there's already a word for truth without belief and that's called "fact".

Otherwise how could novelty ever enter the picture? — Wayfarer

Novelty is relative. As long as you couldn't predict it, it's novel - and you can't predict reality perfectly no matter how deterministic it is. You can predict certain low -complexity events, like the approximate location a bomb will land of you launch it at a particular angle with a particular amount of force, but you can't predict the future of a brain faster than the brain can do something that might surprise you, even if the processes in the brain are effectively macroscopically deterministic.

you just said that physics is 'determined by subjective requirements'..... — Wayfarer

You put those words in quotes as if I literally said them, but I didn't say them.

I'm flummoxed as to why you or anyone would find deteminism beautiful — Wayfarer

Order. In a deterministic system, every event has its place in the system, every event has a clear explanation and follows from the way the system is. In an indeterministic system, there's chaos because "stuff just happens". There's nothing particularly beautiful about "stuff just happening", compared to the beauty of patterns and order. And you don't have to accept that, of course, it's not some kind of scientific fact that that's what beauty is. I'm not telling you you need to believe that, just saying why I do.

I think a surprising amount of physics is based on abstract, apparently-subjective judgements of physicists. What would be a beautiful way for the world to work? Look at how much talk there is of symmetries in modern physics.

I think there's an intuitional gulf between those that think that's a reasonable guiding principle and those that can't see why beauty should have anything to do with how reality operates. I personally can't justify why physics ought to be beautiful, but I feel in my bones that the success of physics ideas influenced by ideas of mathematical beauty are no mistake, no coincidence.

A deterministic world is ordered, logical, "everything in its right place" - indeterminism includes things that happen for no reason, which is inherently disordered and chaotic. It's just not very neat.

there are multiple deterministic interpretations of qm too so we can keep the beauty of determinism anyway.

So why believe it? — Moliere

Because it's beautiful (in a mathematical sense). There's something hideous about random spontaneous things with no ontological explanation just popping into existence.

I believe Einstein was hinting at that when he said God doesn't play dice.

to me, the most disappointing part of the audiobook is her reason for cutting her conversation with Sean Carroll short. It seemed like she was saying "he didn't like my idea, so I didn't want to talk to him any longer". Obviously it's her book, her right to direct the direction of the content, but I think they could have had a much longer, more interesting conversation if she allowed the conversation to explore an alternative viewpoint of consciousness.

she says herself that she's a physicalist multiple times in the audio book. She also says it here.

https://open.spotify.com/episode/4uUFfDc8HYTsdbTengp1A2?si=03MxDbP0SsO8Q1yMUT4gAQ

I guess she means, it's fundamental in a way that's somehow tied to physical things. I don't really know tbh, I'm not her and I don't have her ideas.

The closest I get to fundamental consciousness is thinking that there might be a "thing it's like to be" any time there's a process that's sensitive to changing variables, and that might just be a fundamental truth. But I wouldn't commit to that position, it's at best a coin flip as far as I'm concerned. But if that were true, it would be I think straightforward to see how physical processes would end up with "what it's like to be"ness