Complex Things Have to Grow

Fridman opens with a blunt question: is there intelligent life out there? Dawkins answers that if we accept it exists here, and accept that the number of planets is gigantic — some 10 to the 22 stars have been estimated — then other life, including intelligent life, is highly likely. To deny that, he argues, is to treat what happened on Earth as staggeringly, "ludicrously off the charts" improbable. He does not think the origin of minds is that rare.

He splits the problem into two hard steps. The origin of life itself is probably fairly improbable, and the later climb from living chemistry to intelligence is also hard. Stacked together they are pretty improbable, but not 10-to-the-22 improbable. The most likely first contact, in his view, would be radio rather than a visit. Researchers in SETI have long discussed how to tell a signal from noise; broadcasting prime numbers would be an obvious way for another civilization to say "we are intelligent, we are here."

Fridman wonders whether alien mathematics would even match ours. Dawkins suspects they will have Pythagoras's theorem, because the same physics should yield the same math. He then sticks his neck out: if we ever discover life elsewhere, it will be Darwinian life, running on the non-random survival of randomly generated codes. It would need some kind of genetics, but that genetics would probably not be DNA. Selection, in a general sense, is the part he thinks a biosphere cannot escape.

Fridman notes that Dawkins has long been an outspoken critic of intelligent design. As far as both men can see, evolution is the only scientific theory on offer for how intelligent life developed. No alternative has ever been seriously suggested, and Dawkins suspects none ever will. Fridman teases the danger of saying "never" a hundred years later. Dawkins allows that a future account might look like Einstein's update of Newton — an alteration of the theory, not a replacement of the core idea.

That raises an engineering question. Can humans do what intelligent-design advocates claim a designer did, and build intelligence in machines, or is some evolutionary process the only path? Dawkins is committed to the belief that it is ultimately possible, because there is nothing non-physical in our brains. They work by the laws of physics, so in principle they can be replicated.

In practice, he grants, the job might be very difficult. The only workable route might even be something like an evolutionary process of variation and selection. He would be surprised if that were strictly required, but he notes that machine intelligence has been slower in coming than some of the early pioneers expected.

Is evolution efficient? Dawkins calls it a paradox. On one side it is deplorably wasteful, "fundamentally based on waste." On the other, it produces magnificent results. The design of a soaring bird — an albatross, a vulture, an eagle — is superb; an engineer would be proud to have done it. The same engineer would not be proud of some of the botched jobs evolution has served up, which make sense only once you know their historical origins.

His favorite example is the recurrent laryngeal nerve. It is a cranial nerve whose end organ is the larynx, yet it does not go straight there. It runs down into the chest, loops around an artery, and comes back up again. Dawkins assisted in the dissection of a giraffe that had died in a zoo and watched the nerve whiz past the larynx within an inch, continue many feet into the chest, then return — a ridiculous detour.

The reason is historical. In fish ancestors the equivalent nerve innervated part of the gills, and the most direct path already ran behind that artery. As tetrapods came onto land and necks stretched, jumping the nerve over the artery would have been embryologically costly, while lengthening it a millimeter at a time was cheap. Each tiny increment did not matter until a giraffe made the cheap path absurd.

Any engineer would reject that piece of design. What surprises Dawkins is not that such examples exist, but that there are not more of them. Natural selection tinkers, cleans up early mistakes, and makes the best of a bad job. Fridman adds that we forget how many generations had to die for the finished product. Dawkins agrees: that is the horrible wastefulness of it.

Fridman floats the thought that humans might be only the biological bootloader for artificial general intelligence. Dawkins agrees that if superhuman intelligence is found, it will be artificial. He has no hope that human brains will keep getting larger by ordinary biological evolution. Brain-size swelling has been a dominant trend in the fossil record of the last two or three million years, but that is unlikely to continue, because natural selection would have to keep favoring individuals with the biggest brains — and that is not happening anymore.

For intelligence to keep evolving genetically, the most intelligent individuals would need the most children. Intelligence may buy money, a nicer house, a career, and admiration, but it does not increase your genetic legacy. Meanwhile computers and technology evolve by non-genetic means, by leaps and bounds. Fridman asks about Neuralink and two-way brain-computer interfaces meant to expand the brain's capacity. Dawkins finds the effort interesting but says he is not qualified to judge whether it is promising.

What puzzles him is the existing brain. It is small compared with a computer, and its components are slow compared with electronics, yet it is astonishing what it can do. Imagine packing a machine into a human skull whose switches work as slowly as neurons: something mysterious must be going on that we do not understand. He had just spoken with Roger Penrose, a materialist who still treats consciousness as a deep puzzle, possibly with a quantum-mechanical aspect. Dawkins will not argue with one of the world's greatest physicists. Nobody really knows anything about consciousness, he says, and leaping from that ignorance to a religious explanation is "an act of cowardice."

Fridman turns to *The Selfish Gene* and the gene-centered view that treats organisms as the medium through which genetic software propagates. In that context, what is a meme? Dawkins says it is the cultural equivalent of a gene: that which plays the same role in the transmission of culture and ideas. The word is only useful if something Darwinian is going on. Culture is obviously transmitted; the live question is whether some variants become more numerous and others less so, in a differential, gene-like way.

Memes clearly replicate. The harder claim is whether they replicate in a Darwinian fashion — whether certain ideas spread because they are successful in the meme pool. He pushes further: animal bodies are adapted as machines for propagating genes. Could a human also be modified as a machine for propagating memes? Even anatomy can change by memetic means. Hairstyles, makeup, and circumcision are actual changes in body form that are not genetic, yet they pass down the generations or sideways like a virus.

Once you leave the physical and enter the space of ideas, ideologies, and political systems, the analogy gets more interesting still. Dawkins thinks memes are a bit more than a metaphor. Religious ideas move in two directions: longitudinally down generations, like conventional genetic transmission from grandparent to parent to child, and transversely, like an epidemiology of belief, when a charismatic preacher converts people in a pattern that resembles a virus.

Fridman asks about social networks and viral spread — the modern use of the word Dawkins coined in 1976. When he first used it, the internet did not exist; he was thinking of books, newspapers, radio, and television. Now an idea can leap around the world in all directions instantly. The internet, he says, is a step change in the facility of meme propagation. Fridman likens the 1970s to isolated Galápagos islands that the web suddenly globalized, with ideas breeding, mutating, and facing selection in seconds.

Dawkins thinks it would be a good idea to think in a Darwinian way about this. Ancestors lived in villages and small bands; ideas propagated locally, hopped slowly to a neighboring village, maybe eventually a neighboring continent. Nowadays villages are international. People live in what have been called echo chambers: internet villages whose members may be scattered across the globe yet share terminology, jargon, and enthusiasms. The Flat Earth Society is his example. Members do not live in one place; they find each other and talk the same nonsense to each other. That is a distributed version of the old village.

Darwinian selection here means investigating which ideas spread and which do not. Flat-earthism has a few believers but is not going to spread, because it is obvious nonsense. Other ideas, even if wrong, can spread because they are attractive in some sense. Attractiveness need not be defined the way animals attract each other; all that is needed is that they spread. Truth is one criterion that might help an idea travel, but not the only one. The peacock's tail does not help the peacock survive; it helps pass on genes. A rubbish idea that people do not recognize as rubbish, and find very attractive, will spread the same way.

Fridman has read arguments that in some birds, aspects of beauty serve no evolutionary purpose of their own. Dawkins thinks that view is rather distorted. When Darwin coined sexual selection, he did not require that what females found attractive had to be useful. It was enough that they found it attractive. A trait could be completely useless in the ordinary survival sense and still be highly useful for passing on genes.

Others, starting with Alfred Russel Wallace, the co-discoverer of natural selection, disliked that idea and wanted sexually selected features such as peacocks' tails to be useful in some sense. Both schools are still active and about equally supported: those who follow Darwin in saying attractiveness is enough, and those who follow Wallace in demanding a survival payoff. Dawkins will not pick a camp. He thinks both could be correct in different cases, and both have been made more mathematically sophisticated than when the argument began. Fridman, calling himself a romantic, prefers the Darwinian side — beauty as a powerful force in evolution and in religion.

Will there ever be a time when almost nobody believes in God, and God is no longer part of the moral fabric of society? Dawkins thinks yes, though it may take a very long time. Asked what could do a better job than religion, he answers with truth and reason — truth about the real world, ascertainable not only by rigorous science but by ordinary sensory observation. Individuals are clearly not hopelessly tied to religion, because many already do not believe.

Some writers argue that society needs a figurehead, even a nonexistent one, in order to function. Dawkins has read those books and does not find them that persuasive. The subtext that "you and I are intelligent enough not to believe, but the plebs need it" strikes him as patronizing. He does agree that humility is right, given that 99.99 percent of what there is to know remains shrouded. Humility does not, however, license supernatural fillers when present science falls short. To reach for a non-physical explanation is, in his phrase, to lie down in a cowardly way and invoke magic.

Can scientific ego lead us astray? Dawkins grants that there are arrogant people in every walk of life, scientists included, including some who think we have solved everything. Humility is the proper working stance because it encourages further work. Paradoxically, resorting to a supernatural explanation is also a kind of arrogance: it says that because we do not understand something scientifically, the religious account must be the right one. He does not see believers as dishonest or despicable — many are very nice people. He thinks they are often making a misunderstanding, jumping from "we don't understand" to a supposed alternative that is actually a non-explanation. They are misguided.

Fridman notes that a conspiracy theorist's credo is to question everything, which is also the credo of a good scientist. Dawkins agrees it is too easy to dismiss something by labeling it a conspiracy. Occasionally conspiracies are right, so each should be examined on its merits. Flat-earthism is obvious nonsense and need not detain anyone long; other theories might actually be true.

Fridman, who grew up in the Soviet Union during the space race, raises the claim that humans never went to the Moon. Dawkins can understand why the story has a market, but rigorously proving the negative is hard; you fall back on intuition about who would have to lie and cooperate. The sheer number of people who would have to be in on it, the attention to detail required, and the missing motive make it extremely unlikely. Fridman's own intuition is that most people are good, which makes a vast dishonest plot improbable. Dawkins also wonders why anyone would want to suggest the landings did not happen, given that the physics and the trajectories work.

Why do people believe anyway? Fridman suggests there is something pleasant about pointing out that the emperor has no clothes, about being the one who sees while everyone else is deluded. Dawkins recognizes the same appetite in science: the wish to prove the entire community wrong. Historically, lone geniuses have sometimes been right, but people who think they are lone geniuses much more often are not. You have to judge each case on the evidence. Being a maverick does not make you right.

Religion, he argues, is a different kind of persistence. It is passed down generation to generation, and childhood indoctrination is a very powerful force — which is why Catholicism and Islam go on. Moon-landing, 9/11, and Kennedy assassination theories are not childhood indoctrination. They are dreamed up by somebody who tells somebody else who wants to believe it. He can understand believing parents and teachers when you are tiny and cannot think critically. He does not understand the eagerness to fall in line with a stranger who spins a yarn.

Dawkins has long argued that humans do not need religion to be moral. Where, then, do morals come from? He calls it a very complicated and interesting question. What is clear is that the moral values of civilization change by the decades and the centuries. Twenty-first-century people are quite clearly labeled as such by their morals. There is a spread — some more ruthless, some more conservative, some more liberal — but compared with the 17th, 18th, or even 19th century, the whole Gaussian distribution has shifted. We are much less racist and much less sexist than we used to be.

That shift, he says, has nothing to do with religion. The morals of the Old Testament are Bronze Age morals: human sacrifice, an eye for an eye, petty revenge, killing people for breaking the Sabbath — inconceivable now, and intelligible only as the product of the desert people who made them up. Scripture always reflected the moral distribution of its moment, and cultural memes now outrun the text. Basing ethics on millennia-old books is a bad idea; more to the point, we already do not. If we did, we would still be discriminating against women, practicing racism, and killing homosexuals.

Of course one can cherry-pick Bible verses that conform to modern morality, and Jesus said some pretty nice things. But the standard used to choose which verses to keep is already 21st-century morality. Why not cut out the middleman of the Bible and go straight to that morality? Why values change is a harder question — another example of cultural evolution, though probably for different reasons than technology. Dawkins is not sure the direction has evolutionary value, or that it could not drift any which way. What it is, is progressive in a transitive sense: each step is in the same direction as the previous step, becoming more gentle, more liberal, less violent — more decent by our standards.

He is careful about the word decent. Genghis Khan would not call a society of "weak" members more decent. With hindsight we see a trend toward our present values and call it progress. Whether we could return to Genghis Khan is an open question; Dawkins suspects we will not. Fridman notes the dark appetite for watching torture and suffering as entertainment, now channeled into sports. Both men, Fridman with family history in the Soviet Union and the Holocaust, agree that darker aspects of human nature can creep back. There have been steps backwards; the Holocaust is an obvious one. Take a broad view of history, though, and the direction is the same.

Pamela McCorduck wrote in *Machines Who Think* that the story of AI began with "an ancient wish to forge the gods." Fridman asks whether the old desire to create gods and the modern desire to create intelligent technology are the same impulse. Dawkins grants a link between the ancient wish to explain away mystery and what science does. As for creating new gods, he recalls a facetious paper that treated Google as a deity we pray to and consult like an oracle. He gets the poetic resonance and would not have bothered to make the point that way himself.

He can see a future in which intelligent machines, or intelligent aliens, are so superior that we might as well worship them. That is highly plausible. He insists on a fundamental distinction, though, between a god defined as something very powerful and intelligent, and a God who does not need explaining by a step-by-step process like evolution or engineering design. Even a magnificently smarter alien would not be a god in the important sense that it did not just happen to be there. It must have come about gradually, presumably by Darwinian evolution. Intelligent design, he says, comes into the universe late, as a product of a progressive process. Most of the work is done by the slow-moving program underneath.

The simulation hypothesis seems to offer a programmer one could ask why. Dawkins says the idea is not totally ridiculous, but you still have to explain the programmer. Even if we are in a simulation, that programmer must have evolved; if she lives in a higher simulation, the meta-programmer must have evolved. You cannot escape a gradual, incremental explanation. He first met the idea in Daniel Galouye's novel *Counterfeit World*, in which operators of a simulated world discover they themselves are inside another simulation. Nick Bostrom later revived it as a serious probabilistic argument and thinks it positively likely. Dawkins does not, in his heart of hearts, believe it, but he likes the argument.

Fridman argues that even a simulated cosmos would have to be grown, not snapped into being. Dawkins replies that the programmers must evolve slowly, yet they could still switch a world on instantaneously, complete with fabricated memories. He has borrowed Bostrom's picture to puzzle over the brain: his then 100-year-old mother marveled at a smartphone whose real power lived in the cloud. Perhaps most of what we credit to the skull is similarly elsewhere, with the head as an interface — a guess he contrasts with Penrose's view that consciousness resides in the brain as part of physics.

Fridman raises Donald Hoffman's claim that organisms perceive what helps them survive, not necessarily the underlying physical reality. Dawkins believes a fundamental reality exists, yet he agrees that perception is constructive — we build models of what we see. Richard Gregory's Necker cube is his exhibit: a flat drawing the brain renders as a three-dimensional cube and then flips, because the data fit two cubes equally well. He has felt the same switch meeting a friend who looked unfamiliar until recognition clicked and the stored face was grafted onto the sense data.

Asked for books that shaped his intellectual life, Dawkins names Darwin first, then admits he is ashamed never to have read him cover to cover. Writing in the mid-19th century, Darwin was, in Michael Ghiselin's phrase, working a hundred years ahead of his time; everything except genetics is amazingly right. He would have been astonished by Mendelian genetics and by Watson and Crick, because the digital nature of inheritance clears up what had been a riddle in his day.

He has also learned science from fiction. Galouye's *Dark Universe* is set in perpetual darkness, its people navigating by echolocation, their Christianity-like religion rebuilt around light so that they say "I hear what you mean." He recommends Fred Hoyle's *The Black Cloud* and *A for Andromeda*, plus Carl Sagan's *Contact*. Hoyle and Sagan share a theme: we will not be invaded by physical bodies, as in H.G. Wells, but by radio — coded information that persuades us to build a computer running alien software. The idea resonates with Dawkins because bodies are survival machines for DNA, and in meme space the invading code need not be physical at all.

Scientifically, the meaning of life is the propagation of DNA, but that is not how it feels. Each of us manufactures meaning by setting goals — a book, a match, the riddle of consciousness — using goal-seeking machinery that natural selection installed for DNA's sake. Mortality saddens Dawkins mostly because he will miss what happens next; eternity is frightening only if you are present for it. Before birth we were already dead for billions of years, and as Mark Twain had it, we never suffered the smallest inconvenience. The best way to spend eternity, he says, is under a general anesthetic, which is what death will be. "We are going to die, and that makes us the lucky ones": most possible people never get born at all.