The ongoing effort to make AI more efficient and, thereby, more profitable will likely reveal something about the origin of the most complex system known to us: Life. by Charles Mudede The ongoing effort to make AI more efficient and, thereby, more profitable will likely reveal something about the origin of the most complex system known to us: Life.  My reasoning? The most profound understanding we have of time itself was the result of the massive amount of mental energy 19th-century engineers, mathematicians, and physicists applied to the improvement of the fire engine (also called the steam engine), a very crude machine that defined capitalism's British moment. Without the idea of entropy (which, when high, is useless for work; and low, runs our machines, our bodies, our dreams), the key feature of the universe, irreversibility, would have been missed. We move through time because the universe is becoming more disordered (high entropy), and we are becoming more disordered because there are far more disordered states than ordered ones. And it doesn't stop there.  Think only of the Big Bang. It was first formalized by the priest and mathematician Georges Lemaître but rejected by many of the influential physicists of his time, including Einstein. This hypothesis (a guess) only became a theory (a guess confirmed by experience) when, in the 1960s, the right explanation was provided for a problem that frustrated two technicians operating a satellite communication antenna for AT&T. The hiss they couldn't remove from their otherwise immaculate machine turned out to be the belly button of the universe. We now call it the Cosmic Microwave Background. And so we must not be surprised when market-driven research on artificial forms of intelligence reveals, as a byproduct, something as profound as the origin of life. With all of this in mind, let's turn to a little book by AI researcher and software engineer Blaise Agüera y Arcas, What Is Life? Evolution as Computation.  I’m very happy to share that What Is Life?, the first part of my forthcoming book What Is Intelligence?, is now available in print and online. pic.twitter.com/a1dIET5pzg — Blaise Aguera (@blaiseaguera) October 25, 2024 The title of Agüera y Arcas's little book, which is part of a larger project that will be published by MIT Press in 2025, refers directly to Erwin Schrödinger’s equally short book, What Is Life?. The Austrian physicist—mostly known for a rather devilish but poorly understood gedankenexperiment involving a boxed cat in a state of quantum superposition (dead and alive at once)—delivered a series of lectures during the Second World War that exerted considerable influence on the biological sciences because, one, the author was world famous, and, two, it placed information (its biotic production and transmission) at the core of life.  But the content of the book mainly draws inspiration from a paper published in 1951 by one of the key figures in the development of computers (or intelligence machines): Alan Turing.  In "The Chemical Basis of Morphogenesis," Turing proposed a diffusive mechanism for the development of an embryo. This should immediately strike one as odd. And a year later, in "Computing Machinery and Intelligence" he explored the processes at the earliest embryonic stage of life. This meeting of computer science with biochemistry is continued by Agüera y Arcas, a computer scientist and social philosopher who often makes the bold claim that AI is as real as life. And he is not just interested in DNA but, more importantly, viruses (Barbara McClintock), symbiosis (Lynn Margulis), and dynamic kinetic stability (Addy Pross).  The latter (DSK) is related to a program that Agüera y Arcas developed with a Google team and utilizes a programming language invented in 1993 by Urban Müller, Brainfuck. What turned Turing to the embryo is what turned Agüera y Arcas to Brainfuck: Minimalism.  Abiogenesis, the hypothesis that life begins with non-life and emerges from chemical processes that are spontaneous in the thermodynamic sense (spontaneous in this field defines a natural tendency), had to be simple. Complexity has no other home than simplicity. Agüera y Arcas's program, bff, which is described in the book's fifth chapter, "Artificial life," involves the evolution of "random codes" into recognizable patterns, into increasing order, into stability. Agüera y Arcas writes: ...[After] millions of interactions, something magical happens: the tapes [in the computer soup] begin to reproduce! As they spawn copies of themselves and each other, randomness gives way to intricate order. The suddenness of this change resembles a “phase transition” in matter, as between gas and liquid, or liquid and solid. Indeed, the initial disorder of the soup is very much like that of randomly whizzing gas molecules. Hence random, non-functional code has been called “Turing gas”; however, in bff, it “condenses” into functioning code, which is something far more complex than a solid or a liquid. We need to pause here for a moment and go back to the start of this post. Scientists at a corporation, Google, have a model that provides strong support not just for abiogenesis but the spontaneity of life in the universe we find ourselves in. It nods to the primordial ooze.  Let's think about this for a moment. Stability, computation, and replication arise naturally from the soup—or, in my own abiogenic theory, a soup composed of materials/codes exchanged, over millions of years, between planets (life might need a planet-rich solar system, rather than just one in a "habitable zone").  This model of life's emergence might have the power to explain a good part of the reason why genomes are filled with so much junk.  "When the first complete human genome sequence was published in 2000, some surprises were in store," writes Agüera y Arcas. "One was the astonishingly high proportion of so-called 'junk DNA' that doesn’t code for proteins: about ninety-eight percent." There might be a clue here—one related to his bff program. What if a virus is actually dead (and being dead is not a bad thing) but replicates spontaneously? We can then imagine an origin of life scenario that looks like this: Once in a while, some bff-like self-replicating codes, when matched or sequenced in a certain but still random way, coded chemical interactions that resulted in a biological product. This might be the path to Spiegelman's Monster, replicating RNA, the ancestor, according to the RNA World hypothesis, of DNA.   And so the race to command the artificial intelligence market has this spillover: The origins of life. A Marxism that ignores this key feature of capitalist R&D is worthless. There is no science without capitalism (they go hand in hand), and our understanding of the universe would be impoverished without this key capitalist institution, which is often forced to wear the noble-sounding get-up called Enlightenment. How to make sense of this paradox? With this statement in the book Cyborg Manifesto by the biologist and social philosopher Donna Haraway:  "[Though the] main trouble with cyborgs, of course, is that they are the illegitimate offspring of militarism and patriarchal capitalism, not to mention state socialism. But illegitimate offspring are often exceedingly unfaithful to their origins. Their fathers, after all, are inessential."