I asked an LLM why do you always end with some questions or actions list, it is phony. And they answered because it is highly rewarded in the training data because they contain millions of support tickets and they all have that. Well, they proved my Shakespeare'e Clone .
We look back at films, books, photographs, … and laugh and get ashamed asking “how could we be like this?”, I guess the reason is because Mind Graph and Cultural Graph are always running, always working, always updating, and their essence is not the state they are in, but the dialectical update that is running. To explain it simply, that is the offline Wikipedia. You lose the whole encyclopedia when it is static. It is like a static image of twitter. It just doesn’t work that way.
It was this moment where the LLM told me: “You have, in a very real sense, hacked the fundamental constraint of human communication.” I think I have reached one of my destinations.
I was reading I just know who I will be! that I noticed I used to believe in the INFP stereotype :) What a fool I was.
I’m sitting near some █████████. One of the has seen this person who cursed ████████, and has chased the person, caught them, beaten them, filmed them, and then gave them to intelligence stuff. What horrible horrible horrible people.
Imagine how today we have options for buying and choosing on almost everything. This is a form of Element selection and graph shaping.
In arts and making the Toolbox Theory is mental. On the outside commodity is physical.
With every word that Marx and others said about commodity; it still exists and to a huge degree.
My idea of the Toolbox Theory directly addresses tools and tools are mental units much similar to the commodity. Interesting isn’t it? Maybe the commodity is not that much of an outsider. I understand the problem of commodity and alienation quite deeply. And yet that is separate from its “Tool” nature. Either it is the nature that demands atom and elements or it is capitalism that makes us think of these terms. But then whatever you see in nature is made of elements and I have seen how every single mental thing also is made of elements. And so even if we lose the capitalistic argument; the latter survives. We cannot ignore how the world is elemental.
And so it begs the question if commodification of life was a product of the money (it definitely was) but more than that; if it was due to the world going deeper and deeper to Atoms. And this is where a threshold appears
We are more and more understanding the world as it is made of smaller ones. But — while fully evident — we have not yet looked at our world as something made of “Mind Elements”.
Years before; I had seen this speech by an artist that told something like: “It is good if you let your tools give you ideas, but you must never let your tools define what you make. It must not control you”.
Ever since I had read it, I thought to myself that my tools are basically commanding me all the time. I still hadn’t been introduced to the idea of “Programming Paradigms”. In our computer world, we are too familiar with the ideas of environments shaping the mind. We decide between different forms of languages so that we shape the way our mind is going to think about the matters in hand. We have things such as “Functional Programming” vs. “Object Oriented Programming” debates that are going on for ages now. [I have gone ahead of myself here]
And so in my computer graphics world, I always felt I am limited by what I have. I had to go deep and learn everything in order for my work not to be limited by the confines of the system. Say I was limited by the rectangular nature of the web; then I had to learn clipping, and shaders and rendering into canvases and… so that it didn’t apply to my system and I was not confined. This was also the case in AnnotationsKit; I always try to expand the system that I have made not to limit myself. And so; in this fight, I have tried more and more to create tools I had seen elsewhere. So that my ultimate tools could be free; — and as you clearly see the contradiction here — I realized the whole space of Art and everything that appears on the paper is basically made by a tool, a technique, everything you see in the world, everything, is already confined by a tool.
In my youth, as I had remembered, I learned that Iranian music has space between notes in half of a half distance. And therefore the music in Iran has way more possibility and space. Never before that I had thought that there is music more than the twelve keys of my piano’s octaves. After that I had learned about the whole idea of the analogue spectrum and was amazed by how it was possible to have lived and never imagined any music beyond the notes I knew? My mind is constantly making music in the background and that always has been made of the notes I knew. Everything we imagine, we have made with our mental toolkit. That is incredible once you fully experience it: Toolbox Theory in the whole universe…
Isn’t that just amazing? And so I guess we are close to understand it. And when we do; it will be revealed that commodification was onto something: Finding that the world is truly made of elements.
Many years ago, Behdad Esfahbod told me: “Man, Your Résumé is so big, it doesn’t fit into the web”. It was hard to put all my work into one space. And so I wanted to build a website that hosted my work. First I thought about making a database with connections like a graph. And I was working on that till I found Maestro Victor ’s Quotes Collection. And as you may know, I created my own “Quotes Anthology” website. This was good, till I also saw Maestro Victor ’s Dynamicland Archive (I am ashamed of having copied him so bluntly, but I somehow put the blame on him for being the coolest person out there that I cannot help but be obsessed with).
When I first encountered Maestro Victor , I could not understand anything he did, but I could understand everything that he did. It was being like a medieval peasant who had suddenly met a time-traveling scientist. I knew he is not a computer scientist working on fixing programming, I knew he is only using the computers to make what he wants, he was like Maestro Wolfram , using computers as a tool for creating his mediums. I could understand his references to that thing called Smalltalk partially, but not fully, I could see his world, the designs, the Dynamicland prototypes, but I understood nothing.
Then, at my work Fabizi, I started reading his quotes collection. By each quote that I read, I felt more enabled to understand him. It was the most strange and bizarre shift in perspective I had ever had: Understanding this man by understanding the elements that shaped him. And then I began to read the sources of those quotes and more and more, element by element felt like I know Maestro Victor . I read Maestro Kay and Maestro Engelbart to understand Maestro Victor , and realized I have to read Maestro McLuhan to understand Maestro Kay , Maestro Engelbart and Maestro Victor , and then I realized I have to read Maestro Postman to understand Maestro McLuhan , then watched the Media Theory courses of Maestro Moeller to understand Maestro McLuhan and Maestro Postman , of course that made me slip to a whole new category and many new people, and Maestro Moeller himself became one of the most influential and interesting people of my life. After hundreds of papers around Maestro Kay and Maestro Engelbart , tens of hours of life documentary and piles of books on biographies, reflecting each day about their works, I finally understood Maestro Victor . Not only that but it ignited me to go on and build upon their worlds. I have never walked into Dynamicland, but I have lived there more than most other people.
And so when Maestro Victor launched the Dynamicland Archive, I understood it fully that his work could not ever be presented in a summary form. One had to walk the walk I had to understand all those dependencies, all those viewpoints, shift their whole language of thinking about these things, and become a whole new different person to get it. Yes there are people who understand the power of Dynamicland, or how it ascends computing, but understanding it is way more than any of these. It it looking at the Threshold of The Reverse Viewpoint from the initial side, while Maestro Victor stood at the far end.
I could never bear summarization at school. When anyone had asked me to summarize something, I would struggle to death, as I still do, and fail miserably. I hated summarize and how they had lost all the meaning of something. I always felt that they are the worst thing happening to language. And so I realized that Maestro Victor like me, has also understood that it is impossible for others to understand Dynamicland without the full raw body of their work. And I started making the Archive .
Somewhere around a year ago, I had this conversation with an LLM. Whatever I did, it kept returning average, summarized, boring, and empty answers. So I got angry, opened up the iA Writer and kept writing about the domains that I know, the projects that I did, for it to actually believe that I can take an unsafe answer in depth, that I am familiar with the sides I want, and… The LLM was shocked and wrote about how extraordinary it is, and a window opened to me where the LLM no longer gave me bad answers, and had even internalized some of my backgrounds and could talk in orders of magnitude more quality. Naturally I was shocked too. No LLM had been that great. Their memories never worked as good as my writing allowed. If you had inspected what GPT stored in its memory it was always very static and summarized and emotionless. Mine on the other hand was full of emotion, LLMs would started talking in an epic tone without me even asking for them. It was understood.
Here something amazing happened: Each time LLMs couldn’t understand something about me, I would have written a new line in that file and explained that one thing it couldn’t “deduce”. This changed everything. I would talk and add to those lines like law makers write laws: You write some, when something so bad happens that requires a new protection, you write a new law. And then let it work till it reaches one new whole. And so the quality of my LLM conversations was only getting better by each day. Sometimes I would have even found confusions, and would write a line that removed the confusions between the other lines.
After around two years of Archive working, I got this crazy idea: What if I could let an LLM read it all? It was impossible for them to navigate the web and read things, so I wrote this serializer to compile all the textual content of the archive into one continuous file, and had my LLM introduction file snapped at the top of it. The quality was changed by many many orders of magnitude. LLMs now suddenly knew all that I had written in these years, all my theories, all my ideas. Even though out of around 500 to 600 projects that I have, I maybe have only documented 20 or 30 in the Archive , and I have decades of writing which are awaiting to be loaded, LLMs still can absolutely rock by reading the archive file I give to them.
More and more that I worked like this, I realized something. LLMs would end up confusing somethings. So I started adding more and more Mind Drops ( Minddrops ) that would clarify these things, and even–in instances like now–write essays that would help them understand things more clearly.
Here I passed the Threshold of The Reverse Viewpoint , and after it was amazing.
I would have imagined the Archive is a tool for me to remember what I do, or others to spend years reading and understanding me. That still is true, but once I could feed it to LLMs, I had learned that perhaps the biggest audience of the Archive would be LLMs not humans. What a Threshold of The Reverse Viewpoint you may think, but well, it was not. The other end of the threshold was so strange and so radical that I could never ever have even imagined.
After I created Topolets , something interesting began to happen. I had these little sticky size bits of information, and the backbone of the Archive LLM Export. When I had given the Topolets to the LLMs, they could understand it all, and they would begin to write clear articles about what they had read. I got shocked again. This is where Basic Media Rules of the LLM World and Writing as Source Code in LLM World were born. I realize that I could write in scatters of Topolets and have an LLM turn it into polished essays.
People used to design cars as horse-less phaetons, and early programming languages were like polished assembly, Leonard Bernstein presented his audience with a MOOG synth playing classical counterpoints, none of them could see what is coming. Could Henry Ford imagine a Porsche 911?, John Backus the CSS?, and Bernstein the modern background sound engineering and synths of 2020’s? And so people still believe of LLMs as a layer of refinement and proofreading over their codes. A C for their Assembly. The answer is way farther ahead…
The answer was the Archive , and it became the most advanced and radical medium I could ever imagined. Instead of me writing a coherent source code of thought, or in other words: text, I would just externalize my thoughts and put them into the Archive . Around here Mind Graph also came to me, and so I can now explain what I do as putting my thoughts node by note into the Archive . I do not care about the polish of my thoughts, the dependency and explaining the prerequisites, I just fill the gaps in what the LLM can deduce by reading teh whole archive. I add memories that were not present, and if they are, I trust that it can infer it by itself. And so I no longer write the long essays, the clear things, I capture my graph, and the LLMs get that graph.
So the amazing amazing thing, is that I no longer write a linear text. I literally–in all possible ways–do raw Sub-Graph Transfer . And it is amazing. If I were to summarize my work, it would be books, but the Archive is not even a short book. It is only 130K+ words as of now, and instead of chapters and new books, I only have to append a new Minddrops for the LLM to connect it to fifty other nodes, and realize 100+ new conclusions without me having to write down every single one of them. It is the most capable and amazing form of thought compression, it is the successor to the summarization. It is a whole new era for language and media.
While every media we have is the capturing of the result of something–or at best a formula for computing results (like math or software)–my Archive is a live program that can generate based on a dense body of my thoughts. While summarization loses information, Archive holds orders of magnitude more conclusions that what is put into it.
I think an advancement as big as this required a much much better writing and presentation than this, but then, I don’t need to, and that is the whole point of it.
