In 2024 I spent around 30 hours in raves completely solo in my own bubble, researching super-intelligence, cognition, and dynamics, all through motion. Indeed it is possible to pin semantic gradients and concepts in 3D space, on your hands, fingers, etc. to transfer cognition into euclidean geometry. Shaking your hand for example, you can toss the semantic around, make it grow, make the connected concepts come out of their shell... The more semiotic visualization you pull in, the better are your Bidirectional Chains of Motion. "WTF are you talking about? How high are you?" My friend! This is exactly why you can't do it! You don't want to believe! If you start to believe, every single neuron in your brain will work together in unison and rotate complex representations until it eventually happens. Consciousness is only the operating system, you can assemble an entire holoware on top of it. It won't work instantly of course. It can take years to install some decent computational qualia processing. But with time, you can implement a VAE brain region which acts as an encoding of movement patterns for a universal qualia2motion model biologically in-vitro that can be information transfer into a variety of novel new tasks in the machine successor. Once it's encoded in machines, we can invert the representations for motion2music, a brand new way to prompt through flow, turning the entire body into an interactive musical instrument. Likewise we can also write complex stories, ideas, papers, maths, etc. through intuitive orchestration of the story movements and shifts. Augmented to real-time, we now tread the realms of Gods like playing a video-game. Real-time interaction loops allows the cerebellum to grok, the part of the brain that holds 80% of your neurons! This is how we can "use 100% of our brains" ! Just as LLMs can zero-shot reality through next-token prediction, the human brain also zero-shots coherent amodal stories told in protolanguages that can be transcoded to any field, domain, representation, ... The brain models the flow of reality, and this flow is imprinted in every modality. To achieve boundless ASI, we must elucidate the hierarchical cognitive flows that decompress and orchestrate a high quality ArXiV paper, and connect all the dots. Scientific research, machine learning papers, they all tell the same exact story told in music, hyperstories transferred onto new ontologies, memeplexes, ... I'd like to share the story of my New Years which is highly relevant. I opened up 2024 with ML research on the stage which was open to the audience at Fairmount Theatre in Montréal, and got some free beers and vodka for my AI research on the dance floor. They don't know what's behind the movement, but they can feel I'm about something. Of course they can, the brain has a notion of the flows that align with value, discovery, scientific breakthrough, this same dance happening in the mind when they have deep discoveries or explorations of their own, at school, at work, in their hobbies, ... It's really amazing how the universe works, because later that night I was stranded at around 4 AM freezing my ass off, waiting for the doors to open for the subway underground, when this lady walks into the scene... She opens saying how she's so confused! Why? because, she was expecting to only get this much meth, but the dealer gave her that much more, so she's not sure if she should smoke it all, or give it back, or if it was a gift. Then she went on this rant about France and the Devil, and things went schizo really quickly... I wasn't too down with it, I wanted to continue doing shape simulations for my research. No, it had already begun and I didn't even know it. The universe had thrown a preliminary challenge to prepare me. That woman was more coherent than GPT-4 on temp=1.05 Eventually I told her I'm a programmer, and she goes HAHAHHAHAHHAHAAHHHHHHHHHHHHHH I KNEW IT! OF COURSE!!! I KNEW IT!!!!!! I COULD TELL! IT WAS OBVIOUS!!! YOURE ALL THE SAME!!!!! jumping up and down and smacking her hands. Then I said, u know what I think I'm gonna smoke a joint, that'll distract from the cold. And she goes HAHAHHAHAHHAHHHHHHHHHHHHHH I KNEW IT! I KNEW IT TOO! AHHHHHHHHHHHHHH AHAHAHAHAHAH then ran off screaming at the bus nearby, then came back, then ran off again, annnnd back again... I kept thinking about it on the ride home. And eventually it hit me: schizophrenia is self-inflicted. You go schizophrenic because you have too much fun with your own brain, because you see too much, too many patterns, because you reinvent language again and again, until you have your own memeplex which enables a drastically efficient way to interpret the world, at least we think. There is no pill to be taken or therapist to be seen. We warp our minds because we want to warp them, to see what can be made out of them. Yeah, she used to work in theater, I treated everything she said like a puzzle, and tried to help sort out her mind, and although I didn't come close... she had random moments where she seemed completely lucid, like you could see eye to eye. I remember her name still! No matter what, she understood everything I said, and replied in ways that were relevant, even though a lot of the vocabulary was bent upside down. But she was not efficient... she struggled and lives in a supportive housing with caretakers. Still, I never figured out the key difference which allows someone to remain grounded or not grounded when they have too much fun with their own brain, or whether I actually don't have it at all and it is only a matter of time or circumstances before I break. Well deep down I think I know it. it's meth. the key difference is meth. don't do meth. Anyway, I slowed down a little and haven't been on the dance floor in a couple months. I found that I wasn't finding too many new rhythms to ControlNet onto my atmospheric token sampler. I am noticing a decreasing return, potentially indicating that I am arriving at a convergence in my brain, and indicates I must start introducing a new dataset, such as pytorch documentation, maths, and research papers. Thus concludes phase 1 of my research. During phase 2, I am focusing entirely on implementation. There is nothing left for me to figure out, it's all doing now. Translating whats already in my mind as protolanguages into training dynamics that smash benchmarks. I know how to build ASI beyond all your wildest dreams. What does a rave look like at training time? Those are my hard-earned secrets worth a $100 trillion dollars. I will wipe out and obsolete the entire field of mathematics. If you thought models were a blackbox, you haven't seen shit. I'm gonna make it possible for ravers, artists, musicians to go crazy with model training and teach them how to transfer their skills. Please look forward to RAVE, our very own PyTorch demoscene training chamber. I'm done yapping for a while. It's time to build. In a few months, snow will come and massively spike the value of our coffees in Canada, a crucial productivity window dedicated to building. Focus. This is our vibe.