00:00So this is a great question. I feel like engaging with AI is literally neurophysiologically in the same category as psychedelics. I generally feel like a combination of uncanny hallucination and dreamlike sensations when I even read their Twitter threads on ChatGPT. I feel it disembodies and atrophies my left hemisphere and shoots paranoia into my right hemisphere. So yes, um, what these models are doing is, it's very similar thing to psychedelics, but it's also a very similar thing as what THC is doing, it's a very similar thing as what iterative journaling is doing, it's a very similar thing as what meditation is doing, it's just not very guided or directed. Now it can be, and it can actually be a very powerful integrative tool that is actually very healthy for the growth and development of the individual, but we have to look and explore what integration actually is.

00:49Integration is when two parts of the brain fire off at the same time that don't normally fire off, and because they fire off at the same time they essentially find a pathway between those two things and it causes a restructuring of those two things to grow. When you take a psychedelic, everything fires off at the same time, so you create this one giant massive thought formed of all the semantics and all the patterns that you have ever acquired, and then you come past the membrane of that, you're like, what was that, how do I even name that, what is the thing that I experienced? And it's, well, you experienced all of your patterns at the same time, and it is essentially the entirety, or most of your connectome, and you're seeing most of the map and you're like, I see everything, everything connects. It's like, yes, because everything is connecting, because it's all firing off, because it's all connecting all the parts.

01:36It's not, it's pretty pretty straightforward and pretty logical stuff, it's not magic, it's just that when things fire together afterwards they wire together. So psychedelics are just raising it pretty, uh, arbitrarily, and you have the work of QRI and neural annealing for healing trauma, which is where they're using that to consciously raise the energy and then re-crystallize the structures of the brain as it cools. It's a guided structured process where they're like, you have some sort of trauma, we are trying to address the fact that when the trauma was embedded in your system there's a high level of energy, and therefore to unlearn that trauma you're going to have to come at a similar level of energy to restructure the activation.

02:23So let's talk about the generative models like GPT and DALL-E. The first thing that most people seem to be doing is to be associating these patterns that they normally would never see together, like Darth Vader Elmo Muppet, something like that, or Kermit the Frog as a race car driver, just like putting together a, matching together two patterns which normally are not associated because we just want to see it. But when you associate those in the field of your mind's eye it causes them to actually change the structure of your brain. And so most people are just doing these things, they're just connecting everything. Right, so if we look at something like schizophrenia, the disconnection theorem, it is implying that there are not functional connections between the different parts of the brain through which you can flow down, and that there are parts that become disconnected.

03:13In this case it seems more like we're saying it's like a hyper-connected in an illogical or unuseful way, meaning that the integration has just kind of been thrown at the wall without any particular reason. So this is, comes down to why I'm refusing to build Irises for people, because the root of Cognicism is not AI, the root was always internal integration, the exploration of one's perception of truth over time, and, you know, iteratively refining one's perception of reality, that's what I cared about. And then the Iris became, and I wasn't even called an Iris, it was just, the machine learning was a tool to deal with the more stubborn people who refuse to acknowledge when their old claims no longer aligned with their new ones.

04:02Like if they just explicitly said something that wasn't true, I found that some of the more stubborn people had a hard time just accepting that they had said something, even if you showed it directly into their their face they would just fight it tooth and nail. And that there were other people who really could accept when they had errored. And so I, I realized that something like a scoring function or something basic like a market was never going to work because there was always going to be these very very stubborn people who refused to integrate the knowledge that was the error, I mean they refused to integrate the error of their perception. So very much what you're seeing with all these tools is a very unconscious undirected integration of the connective map, and then we're discovering that it's causing people to have some very interesting behaviors.

04:50Now if you look at me recently, I've been highly integrated, but that's because I recorded the memetics video, which was a reintegration of a theory that I had developed five years ago. When I developed this theory five years ago by writing, you know, the manifesto and working through my thoughts, it integrated my mind into a functional and structural connected, uh, representation that was helping me understand the world better. At the time I had the same type of pushback where it's like, well you're too, you have too much right now, you have too much energy, you're not functioning in the same way, and I don't particularly like that, and so I'm going to give you a lot of negativity. And at the time I was like [ __ ] this [ __ ], I'm very lonely and slipping away. Now I'm like, integration is great, and I have other people around me like Forrest Landry who has expressed that he's functioning in a similar state of consciousness.

05:39He's acknowledged me like, yeah this is kind of lonely, but like if we're on this other side and if we understand this conscious integration and if we can show other people how to do the same, then these tools can be very very healthy and help us grow as individuals. But if instead we look at it as a product from Daddy Corporation to save us, it's just going to harm us more and more and more. These tools cannot be passive. The problem with society is not that we don't have viruses, the problem is that we don't listen to each other, we don't pay attention to each other, the problem is that we're always pushing memes into each other and we don't take the time to see if we really have understood the meaning of the other. So it's like we're always interfacing with the simulation of the other generated about our brain, and that simulation is constructed from the nature of the connectomics.

06:27And so now, you know, because of search you have any random person who doesn't have a very well integrated brain, they can search out anything, and as soon as something creates a connection in their brain and makes a shorter pathway to make cognition easier, they're going to believe that, not because it's rational but because it's relative to their individual representation of knowledge. That just by giving these integrative tools to people who aren't using them in a directed way and are just kind of searching things or asking random questions, most people are going to seek to integrate the furthest part of their mind. It's a natural thing, it's a physical thing. But what you're going to end up with is not with a usefully integrated structure that manifests coherent thoughts.

07:15You're going to end up with a sensation of hallucinogen or a sensation of atrophy. So what I would suggest is, if you're feeling this way by interacting with GPT, that you should start a journal and create a data set that is essentially your fine-tuned model. Right, because by iterating with not just a model that represents global knowledge, right, and instead by integrating with a model that is your knowledge, you see the map before you and you can consciously draw the map to make it more efficient and make it easier to find your way through your own mind over time. [Music] [Applause]

08:03[Music] Thank you. [Music]