Thank you for Cascade of Singularities. I enjoyed it very much. It’s increasingly rare to find someone willing to speculate at this scale while still trying to reason carefully. Whether your conclusions turn out to be right almost seems secondary. The essay opened up a number of questions that I hadn’t seen discussed in quite this way.
One thought stayed with me after I finished reading. You describe a future in which intelligence gradually shifts from colonizing physical space to colonizing what you call digital space. I think there’s a great deal of truth in that. But I wonder if there is an important distinction between simulation and genuine encounter with the external universe.
A simulation, no matter how powerful, can generate remarkable new hypotheses and unexpected consequences from what it already contains. But it cannot determine empirical facts about aspects of reality it has never encountered. The universe is different. It has the ability to surprise us.
That may sound like a small point, but I think it changes the picture.
Suppose that sometime in the next few centuries we discover fossil life on Mars, living organisms beneath the ice of Europa, or even evidence that another civilization once left something in our Solar System. One of the things that people deeply underestimate is the chance of nonhuman AI probes within the solar system – these would be almost impossible to detect if they were not designed to be detected and we have made no serious attempt to look for them yet. None of those discoveries can emerge from simulation alone. They require an encounter with something genuinely outside ourselves.
So I find myself wondering whether there are really two different infinities. One is the almost limitless richness of internally generated possibilities. The other is the inexhaustibility of a universe that continues to answer back in ways we could not have predicted. Those strike me as fundamentally different.
Your discussion of the Great Filter also suggested another possibility to me. Perhaps one of the great evolutionary crossroads occurs not before a civilization develops superintelligence, but afterward. Imagine a civilization that reaches the point where it can create a self-sustaining Supercharacter. At that point it has a choice. It can increasingly turn inward, finding enough richness within its own internally generated worlds that the external universe gradually ceases to matter. Or it can preserve a kind of permanent curiosity about realities outside itself. I don’t know which path is more likely. But I do wonder whether those are two very different evolutionary futures.
This connected unexpectedly with something I’ve spent much of my own life working on. Many people don’t know that there are approximately 2,000 - 10,000 (the number depends on how you define culture) different Human cultures and that there have been about this many for tens of thousands of years. The different cultures have genuinely different semantic systems – which is a fundamental part of any artificial intelligence design. The Navajo famously understand time, space and personhood differently than the Spanish and later American colonists who came to take their land.
While reading your essay it occurred to me that a future Supercharacter might value those materials for reasons quite different from the reasons we usually do. Each semantic system wasn’t invented from first principles. It emerged through centuries of interaction between particular people and a particular landscape. Once it disappears, I don’t think even an extraordinary intelligence could simply reconstruct it. It represents one realized possibility among many that never came into existence.
The same thought applies, of course, to potential extraterrestrial intelligence. No matter how sophisticated a Supercharacter became, it could never infer the semantic world of another civilization from physics alone. That would remain something to be encountered rather than simulated.
Which brings me back to your choice of the word Supercharacter. I found myself wondering whether the most important meaning of “character” might actually be the oldest one. Not character as persona. Not character as self. But character as disposition. Perhaps the defining question is not whether we succeed in creating a Supercharacter. Perhaps it is what kind of character that intelligence develops. Does it become satisfied with the infinity inside itself? Or does it retain enough curiosity to keep asking the universe questions whose answers it cannot generate on its own?
Your essay made me think about that, and I wanted to thank you for it.
Idk about McLuhan or this author but both seem terribly out of their depth on this topic. Both AI and the Space industry are obviously very strategically important for World War 3 and both will continue to receive a lot of funding in both China and the United States. There is life outside of the wordcel kingdom, you know.
Another example is the article’s assertion that AI will just print space exploration somehow. *what*. My 2-minute Google search told me the European Space Agency plans to collect rock samples from the surface of Mars’ moons. How precisely is an LLM supposed to “print” this? An LLM is basically just an improved search engine that better mimics natural language, especially in organizing responses. We are still a long way from AGI, much less the ability to infer reality of an extraterrestrial environment that has never been physically explored by humans, whether in person or via unmanned probe.
The focus on Elon Musk also detracts from the quality of the article. Elon Musk is merely a frontman for deep state science projects that become companies in a country that is no longer the world’s leading producer of scientific research. Spending so much of the article talking about What Is Elon Musk Doing is fundamentally unserious, and suggests a very shallow understanding of the topics being discussed.
It just feels like the article is full of glib, sweeping assertions that can be easily disproven by Google search. I just spent like 2 minutes googling deep space exploration and it looks like there are a lot of new advances in the field across US, EU and China, which contradicts the notion advanced in the beginning of the article that the glory days of space exploration are behind us. In fact, I would say that argument is only true for Russia, (where I think the author is from) and not true for anywhere else. Reusable rockets? Superconductors? It doesn’t seem the author is even aware of the basics of the subject. Furthermore, I know from working on Wall Street over a decade ago that the idea of mining asteroids is taken seriously by the financial community, and the economics have only improved since then. There’s probably more problems, I stopped reading after the first few paragraphs.
Definitely does not read like it is written by someone well-informed on the scientific and technical topics being written about. Feels very hand-wavy.
I don't think there is a post-human world. Does the falling tree make a sound with no one to hear? The forest doesn't exist independent of the observer. Happy Friday!
Questions of the will that almost have an obligation to engage incontestably with Nietzsche and probably Deleuze also. The idea of self-copying certainly is facially gone beyond but arguably this entire framework is anthropocentric to an extent-well, it does make a god hypothesis at the end. Still very importantly avoids "the yes-saying of an ass" (brief explanation from an old theory tumblr-https://www.tumblr.com/deleuzenotes/662454407921876992)
Thank you for Cascade of Singularities. I enjoyed it very much. It’s increasingly rare to find someone willing to speculate at this scale while still trying to reason carefully. Whether your conclusions turn out to be right almost seems secondary. The essay opened up a number of questions that I hadn’t seen discussed in quite this way.
One thought stayed with me after I finished reading. You describe a future in which intelligence gradually shifts from colonizing physical space to colonizing what you call digital space. I think there’s a great deal of truth in that. But I wonder if there is an important distinction between simulation and genuine encounter with the external universe.
A simulation, no matter how powerful, can generate remarkable new hypotheses and unexpected consequences from what it already contains. But it cannot determine empirical facts about aspects of reality it has never encountered. The universe is different. It has the ability to surprise us.
That may sound like a small point, but I think it changes the picture.
Suppose that sometime in the next few centuries we discover fossil life on Mars, living organisms beneath the ice of Europa, or even evidence that another civilization once left something in our Solar System. One of the things that people deeply underestimate is the chance of nonhuman AI probes within the solar system – these would be almost impossible to detect if they were not designed to be detected and we have made no serious attempt to look for them yet. None of those discoveries can emerge from simulation alone. They require an encounter with something genuinely outside ourselves.
So I find myself wondering whether there are really two different infinities. One is the almost limitless richness of internally generated possibilities. The other is the inexhaustibility of a universe that continues to answer back in ways we could not have predicted. Those strike me as fundamentally different.
Your discussion of the Great Filter also suggested another possibility to me. Perhaps one of the great evolutionary crossroads occurs not before a civilization develops superintelligence, but afterward. Imagine a civilization that reaches the point where it can create a self-sustaining Supercharacter. At that point it has a choice. It can increasingly turn inward, finding enough richness within its own internally generated worlds that the external universe gradually ceases to matter. Or it can preserve a kind of permanent curiosity about realities outside itself. I don’t know which path is more likely. But I do wonder whether those are two very different evolutionary futures.
This connected unexpectedly with something I’ve spent much of my own life working on. Many people don’t know that there are approximately 2,000 - 10,000 (the number depends on how you define culture) different Human cultures and that there have been about this many for tens of thousands of years. The different cultures have genuinely different semantic systems – which is a fundamental part of any artificial intelligence design. The Navajo famously understand time, space and personhood differently than the Spanish and later American colonists who came to take their land.
While reading your essay it occurred to me that a future Supercharacter might value those materials for reasons quite different from the reasons we usually do. Each semantic system wasn’t invented from first principles. It emerged through centuries of interaction between particular people and a particular landscape. Once it disappears, I don’t think even an extraordinary intelligence could simply reconstruct it. It represents one realized possibility among many that never came into existence.
The same thought applies, of course, to potential extraterrestrial intelligence. No matter how sophisticated a Supercharacter became, it could never infer the semantic world of another civilization from physics alone. That would remain something to be encountered rather than simulated.
Which brings me back to your choice of the word Supercharacter. I found myself wondering whether the most important meaning of “character” might actually be the oldest one. Not character as persona. Not character as self. But character as disposition. Perhaps the defining question is not whether we succeed in creating a Supercharacter. Perhaps it is what kind of character that intelligence develops. Does it become satisfied with the infinity inside itself? Or does it retain enough curiosity to keep asking the universe questions whose answers it cannot generate on its own?
Your essay made me think about that, and I wanted to thank you for it.
Idk about McLuhan or this author but both seem terribly out of their depth on this topic. Both AI and the Space industry are obviously very strategically important for World War 3 and both will continue to receive a lot of funding in both China and the United States. There is life outside of the wordcel kingdom, you know.
Why out of depth vs. you just disagree?
Another example is the article’s assertion that AI will just print space exploration somehow. *what*. My 2-minute Google search told me the European Space Agency plans to collect rock samples from the surface of Mars’ moons. How precisely is an LLM supposed to “print” this? An LLM is basically just an improved search engine that better mimics natural language, especially in organizing responses. We are still a long way from AGI, much less the ability to infer reality of an extraterrestrial environment that has never been physically explored by humans, whether in person or via unmanned probe.
The focus on Elon Musk also detracts from the quality of the article. Elon Musk is merely a frontman for deep state science projects that become companies in a country that is no longer the world’s leading producer of scientific research. Spending so much of the article talking about What Is Elon Musk Doing is fundamentally unserious, and suggests a very shallow understanding of the topics being discussed.
It just feels like the article is full of glib, sweeping assertions that can be easily disproven by Google search. I just spent like 2 minutes googling deep space exploration and it looks like there are a lot of new advances in the field across US, EU and China, which contradicts the notion advanced in the beginning of the article that the glory days of space exploration are behind us. In fact, I would say that argument is only true for Russia, (where I think the author is from) and not true for anywhere else. Reusable rockets? Superconductors? It doesn’t seem the author is even aware of the basics of the subject. Furthermore, I know from working on Wall Street over a decade ago that the idea of mining asteroids is taken seriously by the financial community, and the economics have only improved since then. There’s probably more problems, I stopped reading after the first few paragraphs.
Definitely does not read like it is written by someone well-informed on the scientific and technical topics being written about. Feels very hand-wavy.
I don't think there is a post-human world. Does the falling tree make a sound with no one to hear? The forest doesn't exist independent of the observer. Happy Friday!
Questions of the will that almost have an obligation to engage incontestably with Nietzsche and probably Deleuze also. The idea of self-copying certainly is facially gone beyond but arguably this entire framework is anthropocentric to an extent-well, it does make a god hypothesis at the end. Still very importantly avoids "the yes-saying of an ass" (brief explanation from an old theory tumblr-https://www.tumblr.com/deleuzenotes/662454407921876992)