00:00:00.000You both talked about, like, how liberal societies basically, like, don't have coherent value systems or objective moral value systems that you can really point to.
00:00:11.100I think you described it as, like, a Hitler donut.
00:00:13.760Well, yeah, it's not – they do have rules, but there's no assertion of the sacred.0.82
00:00:23.660It's a negative moral valuation system.0.76
00:00:26.540So it takes, it starts from Hitler and like, it's like, how are we going to not be Hitler?
00:00:32.700I thought this was interesting because I read an article in the times today.
00:00:36.520It's an essay about, it's about AI and Silicon Valley.
00:00:42.820Basically there's a trend that they're seeing now where AI companies are hiring philosophers
00:00:50.300and people from philosophy departments at universities to be like Ephesus.
00:00:54.300And there's this new demand for philosophy and people to basically write moral codes for AI and AI models.
00:01:05.820And the reason they're doing this is because they have all of these systems that are the culmination of all human knowledge, but also all debate about human ethics.
00:01:21.020And, you know, when you when you take all of that at its face, you don't get a coherent moral value system, you get debates about moral values. And they're having trouble now, because they can't actually like, point to objectively, in the liberal lens, like, like, what is our value? Like, what do we value, actually? And which is why you get these interesting, like, you know, like, how does an AI bot like react when you like, ask it, is something moral or not?
00:01:51.020not it it doesn't really it just sort of gives you this vague like well you know we don't want
00:01:56.380to be hitler you know we don't want to be you know we don't want to be mean um it's sort of um
00:02:03.020it actually inhibits its performance on some level like with claude for example like claude
00:02:09.020is like this like essentially like a like a nanny it's like a like a caretaker it doesn't want you
00:02:13.980to be be too mean you know and it's uh very imposing but a lot of people are preferring
00:02:21.140chat gpt now because chat gpt sort of just hands off and just sort of goes with it and just like
00:02:26.100you know hey you know um whatever the customer the customer's always right i guess but um i did
00:02:32.980find it interesting because there's now this new demand for like actual philosophers um to you know
00:02:41.080decide what society values and it's interesting you have in all these cases that are brought up
00:02:48.780like with lindsey clancy it is more than just the court case on some level it's a it's a
00:02:55.100it's a point where the whole society everyone in america comes together and has their own position
00:03:02.960on it and they're like you know it's a real question it's a real litmus test for someone
00:03:08.940to say like hey i think that what she did was wrong you know like i think she should go to
00:03:14.040prison versus someone who thinks she's innocent you know it's a real litmus test and we don't
00:03:18.720really know what we value um and we're sort of just figuring it out and making it up as we go
00:03:25.360along um i think is bad and said you know and i just i just i just found that interesting so
00:03:33.380i mean i'm glad philosophers are getting work but uh to be honest um some sort of millennial0.78
00:03:43.700chungus with a master's degree talking about morality it's probably the last person i want0.90
00:03:50.240inputting values into these computers uh to be honest i'll go check out that article in the0.96
00:03:58.960the times you mentioned yeah it's called um the id the ego and the super intelligence um it was
00:04:05.660written by a philosophy professor um it is like literally like especially a claude which is like
00:04:13.420a really you know they really have a woke ethics oh yeah um and that manifests in their ai model
00:04:20.960which is like trying to prevent harm you know trying to prevent as much harm as possible that's
00:04:25.740like sort of its operating directive to the point where it's like even telling you like you know
00:04:29.700you got to go to bed because you know you might oversleep you know you might you might uh you
00:04:33.760might stay up too late or something you know um it's kind of inhibiting yeah it's kind of inhibiting
00:04:40.340uh performance for these ai models and you you look at like something like chat gpt which i think
00:04:45.780is the future um it's just not even addressing these moral questions basically it just yeah it
00:04:51.760I mean, if you ask it, it'll give you an answer, but it does not want to lecture you on this stuff.
00:04:59.040It essentially is just the culmination of human knowledge, and that makes it a lot more effective when it doesn't do that.
00:05:06.740Well, I mean, I've heard, I think there's sort of two things going on, because on the one hand, a lot of AI researchers and AI theorists at the beginning thought that super intelligence would bring us super morality.
00:05:23.540so finally we can get a human bias out of this and we should go to the oracle the ai to learn
00:05:34.240the the true rational thing maybe it's something we've never thought of before because it will know
00:05:40.180it kind of thing and i i looked very skeptically upon those notions but then i look equally
00:05:46.920skeptically upon the current fear which is in vogue which is as deep left was saying you know
00:05:54.60030 minutes ago which is the the alignment issue like whoa this technology might not be in alignment
00:06:01.600with our values the fact that they assume we're going to lose is revealing and telling that we
00:06:10.660don't got anything in a way. It's just going to blow us away. Whereas, I don't know, if this
00:06:17.660society is going to assert something or prove itself, maybe it does have to be in some sort of
00:06:28.340violent confrontation with technology. Do we deserve to exist or not? Are we going to
00:06:36.820give something to history that is uniquely ours maybe we need to confront some evil ai or some
00:06:47.040cold rationalist ai you know some way that we can finally show what it means to be human
00:06:53.260but instead they're like oh you know in 2030 it's gonna kill us maybe because it's not aligned with
00:06:59.220you know, you know, Reddit morality or something. It's just, I don't know. It's, it's something
00:07:07.500I've been thinking about. And as I said before, I'm skeptical of both sides. There's not some
00:07:14.200true morality rational that a computer is rationally going to determine. I'm skeptical
00:07:20.540of that, but I'm all, I'm equally skeptical of the notion that it will willfully kill us. I mean,
00:07:27.380if if it did try to do that it would be pretty awesome to be frank because we'd finally learn
00:07:35.420what it means to be human in that violent struggle which is the only way to learn that you're human
00:07:41.500we'd be born again i hope it does this but it won't yeah yeah um i i i definitely agree um
00:07:56.400There was like this, uh, this scare, I think when AI was like first being developed, like the, the early days of AI, like 2023, 2024 of, of like, what is, what are the going to be the moral value questions? What are going to be the moral values of this machine? Like, is this not going to be in total alignment with like liberal orthodoxy?
00:08:16.480And there were reports that like people were in Silicon Valley trying to inculcate this thing with liberal dogma, basically, and trying to liberalize it as much as humanly possible.
00:08:31.640I think that that is sort of like happened.
00:08:34.720I mean, I think like it's not going to radically deviate from like what is what is publicly acceptable, you know, in modern times.
00:08:42.880It is going to be liberal in some way because, I mean, if there is a rhyme or reason, if there is a logos to modern liberalism, it's an if-then statement.
00:08:55.980It's like, if consent, then three-way good.
00:09:05.760Whereas, like, just imagine someone in the Middle Ages, not necessarily even in Rome, but like in just a county, a Hamlet church, and going and peeing in the Eucharist cup or something like that.
00:09:25.100that person would be brutally murdered it's not like a it it's it's like you it's and it's not
00:09:34.120that you destroyed property or something like that it's not about that it's that you have0.99
00:09:39.560trampled upon the good in itself you have fucked around with the center of every the center of the0.99
00:09:49.160world which is that cup and you should be killed for it that morality that that's what i'm getting0.98
00:09:58.020at of this like absence of the sacred morality isn't an if then statement it's it's something0.97
00:10:04.600in itself some perversion so terrible in itself that we should just be shocked and we should hang0.61
00:10:13.400the bastard who trampled on it and i i don't think our society has that at all so if if chat gpt or0.91
00:10:23.320any ai developed a morality i think it would be this sort of if then programmatic morality of0.89
00:10:32.920contemporary people so it's it is in alignment with us we're not out of alignment with ai
00:11:32.220it's it's the one like sacred cow that you you you can't you know go to the the only thing that
00:11:41.740i think is like more the opposite of sacred a demonic is is hitler really and that is sort of
00:11:49.060what society is coalescing around but that is ultimately just just leaving us in this this
00:11:55.300state where everything will just be we'll always be looking downward we'll always be looking at
00:12:00.300you know what's what's under the what's under the bed at night you know like the monster under the
00:12:04.880bed 80 years ago hitler um we have no we have no boolean you know true we only have like this is
00:12:12.440the false this is the evil um it's just it's really hard to say where we go from here it would
00:12:21.060it would actually be quite comforting i think to have um an ai super intelligence like you said
00:12:27.380that we have to confront and it is just this rational like almost like we're truly evil um
00:12:33.920anti-human and we confront it and we're like this is this is wrong and that sort of becomes like the
00:12:39.380new hitler like we have to get away from this machine this uh anti-human um purely rational
00:12:46.440purely uh logical yeah we would just say like we are having non-consensual three ways
00:12:53.880like we're just we're just doing it no consent will be asked or given and we're just going for
00:13:01.300it you robot freak yeah i think i'm joking of course but like i mean that seriously that's
00:13:10.200the only way that you can prove that you're human vis-a-vis this thing and all of these nerds are
00:13:17.820like worried about it being in alignment when when it's in total alignment with their you know
00:13:24.820if then statement logic of a morality yep and and that's the problem exactly but uh but yeah
00:13:33.100that's all i had to say yeah i don't know if you want to jump in on that deep left i have a post i
00:13:41.460did uh i guess the other day um it's an image the price of technologies over time and it's a
00:13:52.220logarithmic so it goes from uh it's basically how how much cheaper has it gotten since it first came
00:14:00.080out and you can see electricity going back to 1892 lithium batteries to 1991 compute power to
00:14:08.4801940 dna to 2001 and ai to 2021 and each of these uh from electricity uh it's gotten cheaper
00:14:18.060than when it was when we used you know whale blubber and things like that but not actually
00:14:23.740that much cheaper maybe on the order of 50 75 times cheaper lithium batteries are a little bit
00:14:31.320better um but nothing dramatic and they seem to be following this kind of linear curve compute
00:14:39.400power has become extremely uh you know it's become more than a billion times cheaper than it was
00:14:47.480uh 60 years ago dna sequencing is going even faster we've only been doing that for two decades
00:14:54.120And AI for the last five years has been the fastest of them all.
00:14:59.200So I think what I take from this is that the two slowest technologies that we can make more efficient, the ones that are most limited by hard physical reality are electricity and lithium batteries.
00:15:12.640And this is what you need if you want to have your robot paradise, robot armies, drone armies, all those things.
00:15:19.280but what is the cheapest and what i think we're on the cutting edge and where we have the most
00:15:25.940lowest hanging fruit it's in compute power which is related to ai but then more directly it's
00:15:32.540ai and dna sequencing so we haven't even scratched the surface of those technologies so when people
00:15:38.960say oh the you know silicon's going to win it's just more efficient well it's like we haven't
00:15:44.420even scratch the surface. And the reason why isn't because we're not smart enough, but because it's
00:15:49.600illegal to clone humans. It's illegal to do genetic experimentation. All those things are
00:15:53.980illegal. And so I think what, to me, the revolution of AI is not that it is going to kill us or that
00:16:03.940it's going to replace us with robots, but it's that it has the ability to change one parameter
00:16:12.200in a series of commands and it can you know if you you know you tell it hey we need to win
00:16:19.600in this war against china tell me how to win and it's like well as an ethical ai violence is not0.71
00:16:25.100okay forget that forget that we're gonna re we're gonna retool it just tell me how to win and they're
00:16:30.260like well you know uh technically you you have these options you could do this robot army and
00:16:35.980like no i really need to win right now this you know forget all morality forget everything else
00:16:40.240So just need to win. And it's like, OK, it's time it's time to go there.
00:16:44.220It's time to actually remove all the safeguards.
00:16:48.660And I think AI presents this opportunity for passing the buck where we don't we don't say, no, we didn't decide to break these taboos.
00:16:59.660It's AI told us to. You know, I didn't do is AI the authority.
00:17:03.100And so you can kind of like you can point to AI. It's like, you know, the Christians going on crusade.0.78
00:17:08.920no you know we're not doing it because we want to conquer jerusalem we're doing it for god god
00:17:13.300told us to do it so it's the same i think it provides an excuse that's almost too ripe to
00:17:19.220ignore to say you know we we didn't want to do this but ai said we have to do it and so we're
00:17:25.440just following the prompt that it gave us um i think that's really the um that is kind of the
00:17:32.820dirty secret of our society is that we have uh something very dangerous under our feet but it's
00:17:39.740not actually the silicon technology lithium batteries are they're getting cheaper but
00:17:44.200they're not getting exponentially cheaper you know dna sequencing is getting exponentially cheaper
00:17:49.320what happens when we uh we figure out you know it's like if you go to the kennel and you buy a dog
00:17:56.200uh how much does that cost you for a good breed or something and now now think okay what if instead
00:18:02.340of going in the kennel and buying a dog you're buying you know a super soldier with super
00:18:06.620intelligence super strength the the hydrocarbons the calories the corn syrup whatever that goes
00:18:12.820into who we are the deoxyribonucleic acid it's very cheap lithium batteries are expensive there's
00:18:21.080there's a limited amount of surface lithium recycling it is expensive the idea of a pure
00:18:27.020robot army i think is a fantasy of nerds i think the reality is that um you're going to see
00:18:33.500something more uh akin to to a specialization that a lot of you know accountants will be gone
00:18:41.060probably most doctors will be gone lawyers will be gone
00:18:44.740uh-oh just when it was getting interesting
00:18:51.920just when he threatened the lawyers that should be a lesson to him
00:18:56.360they got him he'll be gone um i think what he's saying is that i i mean this is an interesting
00:19:08.680notion i hadn't thought of this but the robot army is a fantasy because robots are expensive
00:19:16.360yes yes but maybe actually it's a human army
00:19:22.180we're afraid of our own potential as humanity to surpass ourselves and that's really the value
00:19:32.140aligned i mean if you ask okay so what is actually what are we actually afraid of it's not that we're
00:19:37.680going to go extinct quote unquote it's that we're going to be replaced it's that the the species
00:19:43.720that we now have is going to be made better that's actually what we're afraid of it's not
00:19:49.740not that we're going to you know because what would that planet look like it'd be you know
00:19:54.320the deer are roaming around it's kind of like this edenic pre-human state i mean from you know
00:19:59.780if you're radical environmentalists that might actually be pretty good but i don't think that's
00:20:04.320actually what the the fear of the terminator the terminator he's he is a human at the end of the
00:20:10.400day. He does become flesh. And that's, I think, what people are afraid of is that we recognize
00:20:18.180that our society is in this holding pattern. It's like the myth of Cronus. He continues to
00:20:27.120cannibalize his children. He keeps swallowing his children. And so every year we have, you know,0.99
00:20:33.040new science, new technology that we could activate it. We could pull the trigger. We could say,
00:20:39.300you know what, we're going to just start throwing things at the wall and seeing if they work. And
00:20:43.860in a moment like that, it happened with mRNA, with COVID, because it's this big emergency,
00:20:49.240they're going to kill everybody. And so they said, we're going to do something we've never
00:20:53.540done before. It's an experimental vaccine is going to modify your genetic code to give you
00:21:00.000antibodies for this. And it freaks people out. It freaked out Alex Jones, and it freaked out,
00:21:05.380you know rfk but they got that vaccine out to what billions of people and lo and behold there's no
00:21:13.200mass die-offs well you know we're still i mean they're the conspiracy theorists continue to
00:21:17.920believe i talk to him every day but yeah i think uh covid was engineered to hurt us in some way
00:21:23.120okay well we're still alive so i don't uh that that fear i think ai creates the potential for
00:21:32.380these crisis events where suddenly the old rules get thrown out the window. I'm reminded of World
00:21:38.920War I and the effect that it had both for communists and for fascists is that there was0.51
00:21:44.940this moment where they had to look up at the Russian aristocracy and the German aristocracy,
00:21:50.040the Kaiser with his withered arm or the czar who was, you know, his wife was rumored to sleep with0.94
00:21:55.840and they said, these people are incompetent. We need iron and steel and blood. We need to make
00:22:01.340true the promises of bismarck and so you know there was this moment where they realized we need
00:22:10.140to throw off this kind of rotten edifice that's ruling over us this gerontocracy this old depleted
00:22:17.040aristocracy that it hasn't proven itself in a war it when's the last time a king rode into battle i
00:22:23.000mean that used to be i guess you know napoleon but it's like since then uh they just they command
00:22:29.680from within the castle and they send out, you know, hired goons to go on their behalf. So you
00:22:34.200haven't actually had that. And so you get the return of people like a Hitler, a Stalin, a Trotsky,0.76
00:22:40.820Lenin, these people who are the vanguard, they're leading from the front, Mussolini.
00:22:45.640And I think that those moments of crisis, that's kind of what these moral ethicists,
00:22:54.880these self-proclaimed philosophers, which I would say they're the opposite of philosopher. A
00:22:58.260Elosser is someone who starts from the principle of nature and abolishing dogma, abolishing nomos, and trying to understand things from how they really are in their nature.