!!!!!!!!!!!!!! this is so!!!!! fantastic!!!!!

Wow thanks algorithm! Seriously why does this not have much more views?

yeah, you RUB that grass. Also i really enjoyed your discussion of it even though i'd already seen the videos you mentioned had inspired you to watch it. Sorry for your mental health XD also me here catching strays, being an El Goonish Shive fan from the Early Days XDXDXD i kinda thought nobody else remembered that comic XD

53:40 I'm in no way qualified to say this, but I would consider getting any form of help for your dysphoria pikuach nefesh, which would have allowed you to "violate" Shabbat if it would potentially save a life (I'm pretty sure saving your own life count as pikuach nefesh, but IDK)

Your writing is so sharp! Love every single little quip you throw in

This was an excellently done video and I'm wishing you the best. I'm so sorry you were exposed to this crapfest of a "movie"

RIP Roosterteeth, long live The Regulation Podcast

Haha! I'm the first of the Saph Cult to comment! But yeah, this is amazing for a <24 hour turnaround. I love AMVs that tell a single story along with their snappy editing. Hope Watcher gets to see it, tomorrow or whenever.

It is indeed pretty simple what happened here.

I was pretty much with you until you went off on the racist society narrative. Bias exists in every society and most people (outside of white liberals) have an in-group bias. Power tends to go to the group with the largest majority. Even when people try to correct for bias it just leads to new forms of bias. Like machines most people make their decisions based on limited information about the world and have blind spots which cause them to fill in the blanks with their ideology. Someone tells you something you want to believe you don't question it, if it's an attack on your side there's nuance and anyone who takes it at face value is an idiot being fed lies.

I like the edit. It gives me 2012 or 2013 vibes 😊

Impressive!

Oof, this hurts. Well done.

Came to see how wrong I am, left seeing how right I am.

I re-watched your video from June just to re-sync. I await your next treatise with anticipation. Intelligence is hard to categorise and or quantify, though can be easy to recognise in interaction with others ( spontaneous humour being one indicator ). I suspect we are several orders of compute away from anything approaching humour, irony or even a play on words.... More please soon :)

Congratulations on being one of the only channels to get me on this accursed day.

thank you c:

better than expected.

Best one I've seen all day!

Finally making some sense on this channel!

It was wrong then and its even more wrong now. Shame on you for being so misinformed and so confident in your ignorance.

How did you find the single thing Yudkowsky is correct about to disagree with (not saying his rationalization is correct, I only watched up to 17:45, but his conclusion is obviously correct)

Damn, i just found out about your video, and i must say thank you for your commentary in this issue. I love your matrix references and the concise and clear explanations. Im wrting my bachelor degree and i have this exact topic as my thesis - Ai does not exits. I agree with everything you said and the only thing i would add to this social/economic/technical perspective is the philosophical one. I think from the point of agency, that all „ai” currently being built is nothing more than a wind-up clock, made to do one specific thing. Calling this inteligence is just blurring the term and making researching the brain harder for all the connectionist researchers because of the focus on generating „AGI”. Would you mind if i contact you with some specific questions? Best regards!

!!!!!!! Amazing!

just downloaded sburb, so relatable

Calculus? As a mathematician I can clearly say you're full of *hit! Calculus is composed of integrals & derivatives that are used for solving curved data! Calculus is actually not used much in real life & I can 100% confirm that there are no calculus applications within any machine learning program or "AI"!!!!! Linear Algebra is also not used & very little Algebra is used at all within "AI" programs! The actual mathematics that are being used in today's "AI" is almost entirely just basic matrix multiplication & I should know since I have studied the mathematical properties of many "AI" programs. I hate it when ignorant people just claim the use of Calculus when they don't understand the math behind how something operates, as if it's the most advanced branch of mathematics & everyone will just accept the statement as fact because so few people have the knowledge to falsify their claim.

Watching this makes me feel like what I imagine it feels like to have a stroke

thanks for the great vid! I would watch a MIRI/Yudkowsky deep dive

uncommon take on the Chinese room experiment: is it possible that the system of the person and the ruleset given to him understands Chinese?

AGI (artificial general intelligence) does not exist but AI does. The actual technology of today (ML/DL neural nets) is referred to as AI. It is very narrow intelligence but that’s all.

You: feeling a significant discomfort from holding a balance for a huge amount of time. Us: 2^10 people feeling a small pleasure from seeing you holding it. Excellent moral behavior, I shall subscribe.

If you haven't read it, check out Beggers in Spain - it explores LeGuin-like versus Yudkowsky-ish attempts at making the future

I've seen a couple of comments now that bring up "scope insensitivity" as a reason that Yudkowsky might be correct, so I want to address that real quick. As I said in the video, Claim 3 of Temkin's argument is that a mild discomfort for a whole lot of people would be preferable to excruciating torture for one person, _no matter how many people are affected._ Falling back to scope insensitivity makes the assumption, as Yudkowsky does in his response, that Dust Speck advocates are appealing to non-linear aggregation, such that the total pain of a massive number of people experiencing Dust Specks is less than the pain of one person experiencing Torture. However, that's _not_ what I'm saying. Put simply, I reject the idea that the sum total of pleasures and pains aggregated across the entire testing population is all that matters when evaluating scenarios. You also have to take into account the density, distribution, and modal average of pleasures and pains. The Dust Specks have a very low modal pain to begin with, and are distributed across a massively large population such that the density of pain is very low, and this holds true _no matter how many people are affected._ Adding more people only expands the distribution, and it doesn't impact the density or mode at all, whereas in Torture, the _total_ pain may be less, but the distribution is far lower, and the density is much, _much_ higher, concentrated on a single person - and the modal pain is also much higher, as a single Dust Speck is obviously far, far less painful than Torture. In this way, we can still say that horrific events that impact a lot of people are worse than ones that impact just a few people, as the modal pain is still very high in, say, an earthquake, while also having a wide distribution and low density, but that one person being tortured is worse than any number of people getting dust specks in their eye. (The density of pain also helps explain why getting a million dust specks in your eye all at once is worse than getting one dust speck in your eye, even though both have low modal pain - the distribution is low, and the density is high.)

I like the video. The Ones Who Walk Away from Omelas is probably my favorite short story of all time. I remember crying when I first read it. I was very young when I read it, but I still get still sometimes get teary when thinking about it. There were things I did not like about the video though. I hope I don't come off as antagonistic, because I overall liked the video, and think we need more like it. Firstly, I think you presented The Ones Who Walk away from Omelas as more univocal than it is. Ultimately I draw the same conclusion from the story you do, but I think the story as written sets up a dilemma (among other things, that story has a lot of depth I think). For me the tension is between Omelas feeling like an utopian city that truly is amazing, so amazing we can't even really comprehend it, and the thing about the child, which is so so terrible, and gives a feeling that something is deeply deeply wrong. I think the ending is ambiguous also. Like those who walk away are not treated as heroes. They're written as ordinary people who have some deep conviction and a certain sense of nobility, but who also might be confused. I certainly don't think its supposed to be a one-sided knockdown argument against totalist utilitarianism, which is how I feel you presented it. Secondly, I think you're poisoning the well a little bit too much, and appear too biased and ideologically motivated. Like putting air quotes around every eliezer/rationalism/effective altruism related thing, assuming that these people are just obviously wrong about everything they say. I tend to agree with MIRI type people about AI, and now I feel this video is very hostile towards me, even though I agree with a lot of the points about totalist utillitarianism. Thirdly, I feel like you don't give a fair representation of the arguments eliezer would've given in defence of his argument about dust specs. I think his most honest reponse to people believing torture to be obviously the worse option, would be that people are incapable of reasoning about big numbers. I think that is what he refers to as "scope insensitivity". Basically: Everyone would agree that 100 people being murdered is worse than 1 person being murdered. But if you make numbers big, people stop being able to reason in this way. Like 10^12 people being tortured will to most people seem roughly as bad as 10^14 people being tortured. I don't feel like you addressed this in the video, and that you just assumed the conclusion eliezer drew from the dust spec thought experiment to just be so blatantly wrong as to not be worth engaging with. Like the "Eliezers capitalist" example you gave, is basically the utility monster thought experiment, and it is contentious. Some people will says its obviously awful, others will say it is good, largely for reasons like scope insensitivity and transitivity. So I don't think its good to bring it up, and just conclude that one side is obviously wrong. I liked the style of the video. And the readouts of quotes. Also appreciate holding up the weight for so long. Imagine that would be tiresome in the shoulder. Sorry for the compliment sandwich I realize I just made, I just typed this out in one go.

found you from tumblr and this video is so good. i need to hear you talk about philosophy forever.

My God, people whose whole deal is "I'm actually really smart, you should listen to me because I'm really smart, don't think hard about what I'm saying because trust me, I'm really smart" are the most annoying people on the planet. Yudkowski leaning into "I don't even have a high school degree" would have been more on-brand, because while that doesn't really say much about someone's capabilities it certainly falls into the "I reject the system because I'm a special genius" narrative these guys love to hype. Somehow I knew about all the LessWrong and basilisk nonsense, but didn't know he wrote a lengthy, rationalist Harry Potter fanfic. Also, friend, you should avoid using so aggressive a noise gate in your audio production. Leaving hardware and recording environment the same, I'd recommend either letting a bit of the dry mix through even when the gate is active, slowing the attack and release times, or using a FIR filter to target the noise frequencies you're trying to suppress. None of these is a perfect solution, but they should give a more even sound with less notable humming abruptly popping in and out of silence.

To simplify for myself. Total utilitarianism is a form of reductionism and therefor wrong because it doesn't take the phenomenon emergence into account. Suffering is emergent 1+1=3 Thought provoking video, thanks.🥰

No idea why RU-vid algorithm decided to push me your video but it was enjoyable. If you'd like some feedback I think you should tone down the despise and avoid so clearly taking a side if you're going to hold a balance for the entire video or maybe you should just grab a gavel and act as a judge. For the content of what has been said itself: -I personally don't see the value of the cutoff point (16:10) and comparing pain levels to torture or hindrance... in a fine spectrum, it doesn't matter what word you choose to represent the level of pain as long as they are standing side by side on a large scale... -Also your comparison to covid is pretty disingenuous since the scale of the problem is entirely different, a million people dying against 5 billions having a small hindrance is hardly on the same scale as 1 person suffering to 2^69 small hindrances...

Although Yud is undoubtedly a weird fellow (and "wrong" on many things), I don't think you are really on point with your criticism. First, the guy is clearly neurodivergent. As such one couldn't assume he means something when that's not explicitly what he means. When he writes 3^^^^^3 that is not the same as 2^69 : 2^69 is big indeed but that's only about a thousands more than the number of possibles values of a byte for example. Eleizer's stupid number on the other hand simply doesn't make sense for the human mind. It is simply too big to fathom. And this is the point where Mr Yudkowky's reasoning "fails": as this number has no practical meaning, this argument has no practical application. There isn't 3^^^^^3 humans to experiment dust specks in their eyes and if there were it would completely change stuff. The reason why our morals tells us to choose the dust specks is because we can't truly believe it will affect that many people (and for any practical situation, we are right about that). Your reverse through experiment (Eleizer's Capitalist) has the same flaw: it can only work if you manage to find a way to reach the stupid number of happiness points for a single individual, but the possibility of such a thing is doubtful at best. And that's not a contradiction with total utilitarianism: you could perfectly have a utility value for each and every thing but for some reason a physical limit that makes it harder and harder to gains HP when you already have a lot of them. Actually this is one of the most consensual fact in economics: people who are already happy are harder to make happier than those who are desperate. To make your experiment more relevant you'd need to provide some justification about how you proceed to make this one agent that happy (it is possible actually to frame it in a way that it can't be countered in Eleizer's Yudkowsky's framework).

what the fuck is this 😂

Great video, 2 problems: Firstly I don't think the Chinese room experiment is a fair analogy, it seems like the type of analogy a lot of artists make:"A.I. is just copying and pasting". its not. it does not have a book from which to copy and paste anything. what A.I. does is learn patterns and extrapolate from them. If the guy in the box was able to learn patterns and extrapolate from them, he would be essentially learning chinese, since that is how humans learn languages in the first place, through pattern recognition(and if not chinese, he would be learning SOMETHING). if u still wanna say the man does not know chinese i guess, you re essentially arguing humans arent intelligent either. It's just not that useful of a definition. Secondly you can say that meaning exists outside communication, but then the question becomes how did you get hold of that meaning in the first place? Your eyes and ears had to communicate with your brain in order for you to understand what "water" is. How come the meaning was transferred? I think you can argue it wasn't. that it was made up IN the mind. the color red doesnt exist, its just a wavelength., and RED was made in the brain. Thats fine, but then again, you re arguing that "A.I. cant do x" by defining x so that we can't either. Its a bit pointless, especially considering at one point we will almost definitely be able to create an A.I. that can take all sorts of inputs(sound,image, text) and maybe even more, which humans dont have(ultrasound for example) and u d still be calling that model not intelligent/ saying he does not truly understand reality. Chances are a good enough model is able to create useful meaning out of the data and is able to extrapolate from it. whether you want to call that A.I. or not is up to you. To assume whatever is going on in the brain can't/won't be replicated seems short sighted to me. I don't think there is nothing special about humans, and technology usually needs way less complexity to get the job done compared to naturally evolved counterparts(A plane isnt as sophisticated as a bird, but it sure as hell can fly)

You totalitarianismified the shit out of me in 10 seconds and made me leave

Very interesting. In my head, the question this dude needs to be asked is; "in what context would you find something to be worth suffering and dying for?" rather than "what are you willing to inflict on the alien?" I'm sure if you wanted to get into the weeds you could make many real-life comparisons to profit & cheapness, and how many people have no problem underpaying people to improve their own existence (from the executive level to the consumer level), the needless consumption of meat for pleasure in advanced economies which aligns very well to the analogy of a child-like entity suffering to enhance the lifestyles of those around them.

so you say that moral intuition should be all that is needed to reject a normative claim? does it apply to, say, homosexuality? aversion has no value in any circumstances when determining the value of anything. yes, here I said it - moral intuition is utterly worthless. it's based on cognitive biases, often magnified by thought experiments; in the case of the dust speck argument, it's the inability to imagine very large amounts or intensity of something. so how can you support utilitarianism if you reject moral intuition, then? the answer is: by relying on phenomenological knowledge. my hedonism does not come from moral intuition, as I radically reject it (and as it is rational to do), but rather directly from intimate knowledge of the very nature of mental phenomena. also, a close analysis of moral claims reveals that when a non-hedonist makes a moral claim, they merely express their emotional attitude and turn it into an imaginary, nonexistent property to project it onto some concept. so when you say, for example, that murder is wrong, you project your dissatisfaction with the idea of murder onto murder, and you pretend your dissatisfaction is somehow a property of the event of murder, which is obviously an incoherent stance, since emotions are not properties of external events. what this means is that the sole objects of true ethics are pleasure and suffering, and moral disagreement concerning metaethics arises when people commit a cognitive error when using these two objects of ethics. the implication here is that entities, including trees, rats, pigs, people, planets don't have any intrinsic value. only events, in the form of mental phenomena, can be intrinsically valuable. even if we assume you are correct when you disagree with the claim above, you've got an additional problem: if you rationally analyze your phenomena, you will see they are all that is needed to explain the mind, and that the self doesn't exist. how can you say the self has any value if it doesn't exist? edit: Temkin seems to have invented terms that sound smart in order to make cognitive biases sound not so much like cognitive biases. 'essentially comparative view'? you've just made up a term, what now? what does it change? so you're taking the 'context' into consideration, and what do you consider in that context specifically? what is its object? what is there beyond the sum of pleasure and suffering? well, the emotional reaction to the thought experiment, in other words, moral intuition. that's it.

_The Dispossessed_ is one of my favourite novels of all time. If you have trouble reading it, I recommend the audiobook, read by Tim Treloar. I don't know if that's available for sale where you are, but it really made the book come alive for me.

Not to be rude, but you look very goofy holding that tiny scale the entire video

Le Guin, philosophy and fanfic nerdery in the same video? I think the youtube algorithm finally did something good. This was thoroughly entertaining, keep it up!

Bruh

Chaia, you have this skill of taking things I’ve never thought about, and making me hyper focus on them for a half hour while you give your essay. Absolutely loved the content, and learned something(s). Great video!❤🎉

That was an excellent demonstration of one of the many ways in which Yudkowskyian total utilitarian rationalism is fundamentally a nonsensical mess. I also appreciated that, unlike a lot of people who talk about "The Ones Who Walk Away from Omelas", you emphasised what the actual point is-a critique of the notion that a flourishing society must be founded upon injustice which tacitly underpins a whole lot of how modern industrial society is organised-while using that as a way to tackle what is truly wrong with Yudkowsky's worldview. I'd love to see you do more on the rationalists, as you seem to have their number in a way few outside of those who cover them semi-professionally do.