Podcast Episode

Welcome to the Crazy Wisdom Podcast. This podcast is for you if you have an insane drive to find the truth of things. It's not the good answers that we seek, but the good questions. I interview a range of different guests from many different fields, all with the intention to uncover the simple truths that are hidden in plain sight. Most people don't want to go there. I go there.

My guests go there and you benefit. Please let me know if you enjoy these episodes and as always, subscribe on iTunes, Spotify, or wherever you listen to the podcast. Welcome to the Crazy Wisdom Podcast.

I've got SDMAT and you can find him on Twitter at SDMAT123. And we're going to talk about theological concepts and consciousness as it relates to AI and some of the foundational models and labs and one in particular about how they approach the outputs of the model and how their approach itself. It's sort of this strange linguistic

beast that with the observer stuff, like everything in reality, the observer plays a really important role. Not only the observer, but the interactorable, the thing that interacts with everything. It's a really strange world we're entering and we're going to talk all about strange world. Welcome to the show. It's true. So theological concepts and consciousness,

because I don't think open AI is, I mean, open AI is a bit of a cult as well. They do the same thing that Silicon Valley does, which is very popularized by Teal, which says you've got to be a cult. You've got to be a cult. And they do it, but I guess it's an outbranch of a cult. So I guess it's a schism. There was a schism in 2021 between OpenAI and Anthropic and led to a new cult.

And then there's this cult is definitely a cult. And it's all about the safetyism cult and not to sort of be totally controversial, but safetyism is important, not safetyism. Safetyism is not important safetyism should be avoided. But we're just kind of in this world where this cult, and now they're like a black hole. They're sucking in all the talented machine learning engineers

into this cult, including Andrew Karpathy and all these other things. I just made some controversial statements. Do you have any thoughts on what I just said? Any questions? Any comments? yeah i mean code is is such a rich and loaded word i think um perhaps that the better one for anthropic in particular is religion well what's the difference i mean but that begs the question

what is the difference between religion and a cult right because they're they're the same thing there's no is there a difference well if if you like them you call them religion but i mean cold cold has connotations of uh people in robes dancing around the fire and there's been a lot of that in discussions and there is some truth to that but i think it's

really interesting to look at anthropic as a religion you know this this isn't business it's theology yeah and you mentioned that they're attracting really talented people well a huge part of that is their their faith i mean they're not just there for the money they're there for their belief. Yeah. But what is the belief structure that unites machine learning engineers

that, I mean, I was going to say that doesn't sort of attract me at all. It makes me kind of run the other direction. It doesn't, like, there's something about both religions and cults. And when I say cult, I actually didn't mean that in terms of negative cults. I do think that there are positive cults and negative cults, a lot of the positive cults turned into the world religions,

but they started off as like Christianity started off as a schism within Judaism that just blew up. And in the first hundred years, it was purely inter-Jewish. It was not something the Romans were thinking about at all, but then it slowly crept up. And so that was its positive cult state. some people would say it was a negative cult, but, but so

anthropic as a religion is really interesting. But what is it about, why is it drawing so many of these machine learning engineers? Like why, why do the machine learning engineers have this belief structure and this faith structure that seems completely different

from the rest of the world? So it is, it is different from the rest of the world. There's there's a specifically silicon valley history here yeah are you familiar with um less wrong and the effect of altruism movement uh yes i am familiar with them i've done an interesting

podcast with caitlin schuler uh she broke down what i think we started talking about which was on twitter which was the um harry potter fan fiction that elizer yudkowski uh built his whole philosophy on right is that where it is yeah yeah so personally I mean I've lived in Bay Area and was somewhat on the fringe of the less wrong movement and at the time that that

was really a radical innovative thinking in a lot of ways what year is this oh that was middle of decade okay so 2015 around interesting yeah yeah yeah so at that time obviously ai was was much more of a speculative um concept but people were thinking forward and the the insight that aliezer has is that um there are ways of modeling what ai can do and what what it will be that have a

certain internal logical consistency and it just so happens that if you play out that that line of thought it takes the form of this um almost apocalyptic outline of the future that has a distinctly religious um character to it in that you have effectively um the devil you you have promises of redemption um you have even have uh demons of sorts i mean the the one that gets talked

about a lot is the basilisk so um what but they don't do they does does you tokowski or to do do any of the other people within this realm ever talk about it in the in the framework work that you just talked about do they ever call it a religion no no i mean um elias is very very clear on this um but when you look at it from the outside it certainly looks like a cult yeah as in both

of senses of the word. And what's really interesting with AI is that those intellectual lines of reasoning