This post is about a paper of mine on accelerating change, the Fermi paradox, and astrosociology, called the transcension hypothesis: sufficiently advanced civilizations may invariably leave our universe, and implications for METI and SETI, published in Acta Astronautica in 2012. In a nutshell, the hypothesis predicts constrained transcension of intelligence from the universe, rather than expansion (colonization) within the universe by intelligence, wherever it arises in the universe.
My paper makes a series of assumptions/proposals about the nature and future of intelligent life in our universe. Most of these key assumptions may need to to be correct, in some fashion or another, for the hypothesis itself to be correct. A few colleagues have asked me to summarize these key assumptions in one place, so here’s my current list. This list is a good way to get a quick summary of the hypothesis as well.
Here are the key assumptions of the transcension hypothesis, as I presently see them:
- Intelligent life, on Earth and elsewhere in our universe, is not only evolving (diversifying, experimenting), but also developing (converging toward a particular set of future destinations, in form and function), in a manner in some ways similar to biological development. In other words, all civilizations in our universe are “evo devo” both evolutionary and developmental. The phenomenon of convergent evolution tells us a lot about the way development may work on planetary scales. A kind of cosmic convergent evolution (universal development) must also exist at universal scales.
- The leading edge of intelligence always migrates its brains and bodies into increasingly dense, productive, miniaturized, accelerated, and efficient scales of Space, Time, Energy, and Matter (what I call STEM compression), because this is the best strategy to become the dominant local intelligence, and because the special physics of our universe allows this continual migration into “nanospace“. Human brains, both individually and collectively, are the most STEM-compressed higher computational systems on Earth at present, but are just now starting to get beat at the production of intelligence by deep learning computers, which are even more STEM-compressed in certain kinds of computation than biological neurons. Fortunately, accelerating STEM-compression of both human civilization and our leading technologies is stepwise testable, as argued in my paper.
- The acceleration of STEM compression must eventually stop, at structures analogous to black holes, which in current theories appear to be the most computationally accelerated and efficient entities in the known universe. Fortunately, this “black hole destiny” for civilization seems testable via the search for extraterrestrial intelligence (SETI), as argued in my paper.
- A civilization whose intelligence structures are compressed to scales far below the nanoscale may well be capable of creating or entering black-hole-like environments without their informational nature being destroyed. There are 25 orders of magnitude in size between atoms and the Planck scale. This is almost as large a size range as the 30 orders of magnitude presently inhabited by life on Earth. We simply don’t know yet whether intelligence can exist at those small scales. My bet is that it can, and that STEM compression drives leading universal intelligence there, as the fastest way to generate further intelligence, with the least need for local resources.
- Due to nature of general relativity, extreme gravitational time dilation (or from the black hole’s perspective, “time compression”) occurs very near the surface (event horizon) of black holes. They act as as instantaneous forward time travel devices, for any civilization able to arbitrarily closely approach the surface of a black hole without destroying itself. In other words, near-black-hole-density entities can meet and merge, effectively instantaneously, from their perspective, with all other civilizations in our gravity well that also turn themselves into such very dense objects. Black-hole-like conditions are thus gateways to instantly meeting and merging with other civilizations in their gravity well, as soon as they “transcend” to black-hole-like conditions. Our gravity well includes the Milky Way and Andromeda galaxies, each of which may contain millions of intelligent civilizations. Perhaps the vast majority of black holes in these galaxies (billions?) are unintelligent collapsed stars. But if the transcension hypothesis holds, some smaller number (millions?) are also a product of intelligent civilizations. In “normal” universal time, galactic black holes are predicted to merge some tens to hundreds of billions of years from now, as our universe dies. But from each black hole’s reference frame (whether classical or “intelligent”), this merger happens near-instantaneously. We can think of black holes as shortcuts through spacetime, just like quantum computers are shortcuts through spacetime. (Quantum physics and black holes may even be unified, or at least connected, in a future theory of quantum gravity). If some type of hyperspace, extradimensionality, or wormhole-like physics is possible, there might also be ways of future humanity instantaneously meeting civilizations beyond these two galaxies (most of the universe appears to be accelerating away from us, due to dark energy). But such exotic physics is not necessary for the transcension hypothesis to hold for the two galaxies in our gravity well (and likewise for all other intelligent civilizations in their local gravity wells). For those meetings and mergers, all we need is gravity and time. Standard relativity predicts that if we can create and survive in black-hole-like densities, and if our galaxies are life and intelligence-fecund, as many astrobiologists think, then we will meet and merge with potentially millions of civilizations instantanously, from our reference frame. In other words, our universe appears to have both “transcension physics” and massive parallelism of intelligence experiments built into its topology and large scale structure, if we ask ourselves what the rest of the universe does if we and other nonlocal intelligences become black-hole-dense objects.
- If we live in not only a developmental universe, but an evolutionary one, each local universal civilization can never be God-like, but must instead be computationally incomplete, an evolutionary “experiment” with its own own unique discoveries and views on the meaning and purpose of life. Thus each civilization, no matter how advanced, would be expected to have useful computational differences, and be able to learn useful things, from every other civilization. In such a universe, we would greatly value communication, assuming that we could trust the other advanced civilizations that we might communicate with.
- If not only intelligence, but also immunity (stability, antifragility) and morality grow in leading intelligences in our universe, in rough proportion to their complexity, in other words, if these three life-critical systems are each not only evolutionary, but also developmental, and thus their emergent form and function is at least partly encoded in the “genes” (initial conditions, laws, and environmental constraints) of the system itself, then we can predict that more advanced intelligences, including our coming deep learning computers, will be not only more intelligent, but also more immune and moral than we are today. This idea is called developmental immunity and developmental morality, and I explore it in my paper, Evo Devo Universe? (2008). If these developmental processes exist, they tell us something about the nature of postbiological life. Such life, and its civilization, is going to be a whole lot more collaboration-oriented, intelligence-oriented, immune, and moral than we are today.
- In such a universe, a moral prime directive must emerge, a directive to keep each local civilization evolving in a way that maximizes its uniqueness and adaptiveness prior to transcension. That means one-way messaging (powerful METI beacons), self-replicating probes able to interact with less advanced civilizations, and any other kind of galactic colonization would both be ethically prohibited by postbiological life, due to the great reduction in evolutionary diversity that would occur. Wherever it happened, we would meet informational clones of ourselves after transcension, a most undesirable outcome. In biology, evolution keeps clonality a very rare outcome, due to the diversity and adaptiveness cost that it levies on the progeny. In such an environment, any future biological humans that wanted to continue to colonize the stars would be prevented from doing so, by much more ethical and universe-oriented postbiological intelligences. That is assuming biological organisms even continue to be around after postbiological life emerges. Due to STEM compression, their status as biologicals would likely be vanishing short, once they invent technology capable of colonization. It seems much more likely that biology develops into postbiology, relatively soon (just a few centuries, if that) after digital computers emerge, everywhere in the universe. This outcome also seems likely to be testable via future information theory and SETI, as I argue in my paper.
- Some physicists, most notably Lee Smolin in his hypothesis of cosmological natural selection, propose that black holes may be “seeds” or “replicators” for new universes. That gives us a clue to what we might do after we meet up with other cosmic intelligences. We would likely compare and contrast what we’ve learned, and then seek to make a better and more adaptive universe (or universes) in the next replication. Current physics and computation theory suggest that our universe, though vast, is both finite and computationally incomplete. It may have gained its current amazing levels of internal complexity in the same way life on Earth got its amazing living complexity, via evolutionary and developmental (“evo devo“) self-organization, through many past replications, in some kind of selection environment, a “multiverse” or “hyperverse.”
- If all of this is roughly correct, our future isn’t outer space, it’s “inner space.” Both the inner space of black-hole like domains, and the inner space of increasingly virtual and computational domains. That’s why the growth of virtual reality, in today’s computers, heralds much more than just better entertainment experiences. Combined with the growth of machine learning, virtual reality is going to become the thinking, imagination, and simulation space for tomorrows postbiological life. Virtual space is where intelligent machines will figure out what they want to do in physical space, just as our own simulating brains are humanity’s virtual reality. And just like humans have have done as our civilization has developed, future machines will do more and more internalization, or thinking in virtual space, and less and less external acting, in physical space, the more intelligent they get. This internalization process has a name, it’s called dematerialization (both economic dematerialization and product and process dematerialization), the substitution of information and computation for physical products, processes, and behaviors. The futurist Buckminster Fuller called this universal process ephemeralization, an idea that contains dematerialization, but not densification. If the transcension hypothesis is true, the ultimate “destiny” of universal civilizations is both “density” (black hole like status), and “dematerialization” (becoming ever more informational or virtual, over time).
As Fermi paradox scholar Stephen Webb says at his blog, this is quite a lot of “ifs!” Disproving any of these assumptions would be a good way to start knocking aspects of the transcension hypothesis out of contention. We would learn a lot about ourselves and the universe in the process, so I really hope that each of these gets challenged in coming years, as the hypothesis gets further exposure and critique.
Perhaps the strangest and hardest-to-believe part of the transcension hypothesis, for many, is its first and ninth key assumption (the two work together), the idea of universal development. The most amazing and odds-defying thing I know of is the process of biological development. Most people don’t think much about how strange and wonderful biological development is, a process guided by a handful of genes in our genome, and they don’t learn much about biological development in college. In many ways, development is even more surprising than evolution, which is process that requires development at the organismic level, and which ignores the possibility of development at the planetary and universal level. But if our universe replicates, and exists in some kind of larger selection environment then not only evolution, but development may very well occur at the universal level. Development in biological organisms is a form of future-specific selection that is far more constrained than what we call natural selection, and if something similar is happening on a universal scale as well, then certain aspects of the future of complex systems are statistically highly biased to converge on particular destinations, and science has a lot of growing up still to do.
The paper’s second key assumption, STEM compression is more palatable to most people, in my experience, and may turn out to be the most enduring contribution of the paper, even if the rest of the hypothesis is eventually invalidated. If you’ve heard of nanotechnology, you know that life’s leading edge today, humanity, is doing everything it can to move our complexity and computation down the smallest scales we can. We have been very successful at this shrinking over the last several hundred years, and our ability to miniaturize and control processes at both atomic and subatomic scales is growing exponentially. In fact, human brains themselves are already vastly denser, more efficient, and more miniaturized computational devices than any living thing that has gone before them. But they are positively gargantuan compared to the intelligent computing devices that are coming next.
Stephen Webb is the author of Where is Everybody?, 2015, a book that offers seventy-five possible solutions to the Fermi Paradox, including the transcension hypothesis. Webb did a great job condensing the hypothesis into three pages in his book. His 2002 edition didn’t include it, as I published my first paper on the transcension hypothesis in mid-2002. Speculation on the Fermi paradox has grown considerably in the last two decades, as it becomes increasingly obvious that we live in a universe likely teeming with Earth-like planets, and also with intelligent, curious life.
Extrapolating todays progress in nanotechnology, any one of these civilizations could send out self-replicating nanotech across our entire galaxy within 100 million years (a reasonable estimate with conventional physics), and beam what it finds out to the rest of the universe (or alternatively, just back to the originating civilization), creating a Galactic Internet, and making our universe as information-transparent as our planet is becoming today. So if Earth-like planets likely emerged in our galaxy at least a billion years before ours did, as several astrobiologists have predicted, then why don’t we see any signs of this Galactic Internet today? Or of past alien visitation, probes, and megastructure beacons near Earth? Or of intelligent structures anywhere in the night sky? In other words, Where is Everybody? That’s the Fermi paradox.
At Webb’s blog, he charitably says the transcension hypothesis is “one of the most intriguing” possible solutions that he has seen. He also kindly observes that “Unlike so many “solutions” to the Fermi paradox, this one offers avenues for further research.” It certainly does, which is why I hope it continues to gain scrutiny and critique.
A few scholars are now citing the transcension hypothesis in their academic papers on accelerating change and the Fermi paradox, including Sandberg 2010, Flores Martinez 2014, and Conway Morris 2016. I am hoping that trend continues. The more attention it gets, the more critique it will get.
Fortunately I think each of the key assumptions outlined above are testable, though some are obviously more testable than others in today’s early stages of astrophysical theory, SETI ability, information, complexity, and evo devo theory, and simulation capacity. If anyone is doing work that might shed light on any of these assumptions, I would love to hear of it.
You can find my paper here: The Transcension Hypothesis, 2012. See also this fun 2 minute YouTube video of the hypothesis, by the inspiring futurist Jason Silva and Kathleen Lakey, which has raised its visibility in recent years.
You can find an overview of the evo devo (evolutionary and developmental) universe hypothesis in my chapter-length article, Evo Devo Universe? A Framework for Speculations on Cosmic Culture, 2008.
Comments? Critiques? Feedback is always appreciated, thanks.