Evolution, Design and the Fermi Paradox

What is the time scale for evolution of complex organisms such as ourselves? On Earth complex life evolved in about 5 billion years (5 Gyr), but one can make an argument that we were probably lucky and that the typical time scale T under similar circumstances is much longer.

There is an interesting coincidence at work: 5 Gyr is remarkably close to the 10 Gyr lifetime of main sequence stars (and to the 14 Gyr age of the universe). This is unexpected, as evolution proceeds by molecular processes and natural selection among complex organisms, whereas stellar lifetimes are determined by nuclear physics.

If T were much smaller than 5 Gyr then it would be improbable for evolution to have been so slow on Earth.

It seems more plausible that T is much larger than 5 Gyr, in which case we were lucky, in a sense I will explain. Inflationary cosmology predicts a very large universe (much larger than what is currently visible to us), so that complex life is likely to exist somewhere in the universe. Conditioning on our own existence (a use of the weak anthropic principle), we should not be surprised to find ourselves lucky -- the few Earth-like planets that manage to evolve life must do so before their suns die. Intelligent beings, while not likely to evolve on any particular Earth-like planet, are likely to observe an evolutionary history that took place over a fraction of 10 Gyr.

Why should T be so large? At present we are unable to make quantitative estimates for the rate of evolution from first principles. It is entirely possible that certain evolutionary steps were highly improbable, such as the appearance of the first self-replicating complex molecules. One can also imagine abstract fitness surfaces with local maxima that trap the system for exponentially long periods of time.

I would not be surprised to find that T is exponentially larger than 5 Gyr. Godel went so far as to propose: "... a mathematical theorem to the effect that the formation within geological times of a human body by the laws of physics (or any other law of a similar nature) starting from a random distribution of the elementary particles and the field, is about as unlikely as the separation by chance of the atmosphere into its components."

The framework described above makes the following predictions:

1. The overwhelming majority of Earth-like planets are devoid of life, thereby resolving the Fermi Paradox.

2. Improved understanding of evolution will uncover highly improbable steps -- that is, improbable even over billions (or perhaps 10^100 !) of years. The fact that life on Earth climbed these steps might suggest intelligent design or divine intervention, but is better explained by the anthropic principle.

See related post evolutionary time scales.

Note added: This idea came to me after reading some discussion of ID and the question of improbable steps in evolution. (Here improbable means, for example, that even in an Earth-size population the multiple simultaneous mutations required jump across a particular fitness valley are unlikely to occur in 5 Gyr given the known mutation rate.) It occurred to me that under certain assumptions what might appear to be ID could actually be due to selection bias -- not taking into account the possibility that complex life is rare even on Earth-like planets. It turns out that the idea is not new -- it goes all the way back to Brandon Carter's 1983 paper which (I believe) coined the term "anthropic principle"! See John D. Barrow and Frank J. Tipler, The Anthropic Cosmological Principle, p. 557 for more. Also see this 1998 paper by Robin Hanson (thanks to a commenter for pointing it out). If one assumes many improbable required steps, one can deduce an upper estimate on the remaining time over which favorable conditions can persist on Earth. Note it appears the remaining lifetime of the sun is less than 1 Gyr, so the coincidence is tighter than I suggested in the original post. As the number of required improbable steps increases the likelihood that we would evolve just before time runs out (favorable conditions end) becomes very high -- this is quantified in the two references above.

The point which I have (still) not seen discussed much is that biologists need not be so defensive about improbable steps in evolution. I sense an almost reflexive -- i.e., prior-driven -- response to any claims of improbability. (On the other hand, perhaps biologists just know more about the details: the claim would be that the historical record does not resemble a typical one that would be generated by a chain of improbable events. I think this question requires further study.) But the main takeaway from this analysis is that improbability does not imply design or intervention.

Life on moduli space?

New paper! Although I'm skeptical about the utility of the anthropic principle (see previous discussion on this blog), I couldn't resist pondering a basic question raised by the large number of string vacua, the vast majority of which seem to be supersymmetric.

Is there something special about exact supersymmetry that precludes complex life? It's a simple question, but difficult to answer.

Life on moduli space?

http://arxiv.org/abs/0908.0943

While the number of landscape vacua in string theory is vast, the number of supermoduli vacua which lead to distinct low energy physics is even larger, perhaps infinitely so. From the anthropic perspective it is therefore important to understand whether complex life is possible on moduli space -- i.e., in low energy effective theories with 1. exact supersymmetry and 2. some massless multiplets (moduli). Unless life is essentially impossible on moduli space as a consequence of these characteristics, anthropic reasoning in string theory suggests that the overwhelming majority of sentient beings would observe 1-2. We investigate whether 1 and 2 are by themselves automatically inimical to life and conclude, tentatively, that they are not. In particular, we describe moduli scenarios in which complex life seems possible.

From the paper:

Assuming our current understanding of string theory is correct, the number of distinct vacua with unbroken supersymmetry and exact low-energy moduli (supermoduli) is infinitely larger even than the vast number of string landscape vacua in which supersymmetry is broken and the cosmological constant nonzero \cite{BDG,DK}. For example, in Calabi-Yau compactifications, the continuous parameters determining the shape of the compact space are themselves moduli and result in an infinite set of physically distinct vacua. Indeed, the highly supersymmetric vacua may be on stronger theoretical footing than their non-supersymmetric counterparts \cite{BDG}.

If complex life is possible on even a tiny fraction of points on supermoduli space, it would be difficult to understand, within an anthropic framework, why we do not ourselves observe unbroken supersymmetry and massless moduli fields.

There is thus ample motivation to investigate whether complex life can exist on moduli space -- specifically, in low energy effective theories with 1. exact supersymmetry and 2. some massless multiplets (moduli).

... Below we list some minimal requirements for complex life. In fact, we do not know whether any of these conditions are necessary or sufficient for life, although it seems they are more likely to be necessary (especially B.) than sufficient. These requirements primarily place constraints on low energy physics. As we discuss below, they do not seem to exclude moduli vacua, at least not in any obvious way.

A. structure formation

B. deviation from thermal equilibrium (long lived sources of free energy)

C. stable matter, complex chemistry

Because inflationary dynamics are typically determined by high energy physics, it seems reasonable to assume that the specific properties of any inflationary epoch (including the spectrum of density perturbations) are independent of the low energy properties of a particular vacuum. Therefore, the requirement of an inflationary epoch neither favors nor disfavors properties 1-2. ...

Small (anthropic?) world

My former PhD advisor gives an interdisciplinary lunch talk at the Berkeley faculty club and gets immediately blogged by Brad Delong!

More on the anthropic principle from this blog. Call me a skeptic ;-)

Lawrence Hall

Lawrence Hall is the name of a building (the Lawrence Hall of Science) and a professor (Lawrence Hall of Physics). The second came to the Berkeley Monday Faculty Lunch Forum to argue that there is empirical content to the Anthropic Cosmological Principle.

What is this principle? Put it this way. Suppose somebody asks you why the universe is pervaded by an 80-20 nitrogen-oxygen gas mixture, or why it is 300K outside. The answer is that the universe isn't like that but that where you are is like that because if you wet surrounded by chlorine gas or in a place where it is 400K you--and all life like you--would be dead. Our confidence in these "anthropic" explanations is strong because we can point to places we know of that lack the 80-20 atmosphere--the asteroid belt--and places where it is not a shirtsleeve 300K--Cambridge, Massachusetts.

Can this anthropic principle be applied to more fundamental issues? Can we say that the mass of the d quark is what it is because if it were 20% lower than the neutron would be absolutely stable and there would be no stars? We cannot see any places in the multiverse where the mass of the down quark is lower, but our predecessors did not know about the asteroid belt and other places lacking an 80-20 atmosphere, and the anthropic explanation for why there is oxygen around for us to breathe was just as valid then for them as it is for us. Is it doing science to use this anthropic principle--or is it just meaningless and tautological? After all, pretty much everything in the universe has to be the way that it is for there to be a physicist with the same name as a building talking in the Seaborg Room of the Berkeley Faculty Club Monday at lunchtime--start with Lawrence Hall as your premise, and you have "explained" everything, in some sense.

Lawrence Hall thinks that there is empirical content, and his argument goes like this...

Living in a simulation

This NY Times article discusses the possibility that we live in a simulation. It's an old idea, but one you have to take seriously if you believe in a technological singularity.

I discussed this in an earlier post, string landscape, AI and virtual worlds, in which I decried the latest anthropic fad in theoretical physics:

...let me now turn to a reductio ad absurdum of anthropic arguments.

Let R = the ratio of number of artificially intelligent virtual beings to the number of "biological" beings (humans). The virtual beings are likely to occupy the increasingly complex virtual worlds created in computer games, like Grand Theft Auto or World of Warcraft (WOW will earn revenues of a billion dollars this year and has millions of players). In the figure below I have plotted the likely behavior of R with time. Currently R is zero, but it seems plausible that it will eventually soar to infinity. (See previous posts on the Singularity.)

If R goes to infinity, we are overwhelmingly likely to be living in a simulation...

Think of the ratio of orcs, goblins, pimps, superheroes and other intelligent game characters to actual player characters in any MMORPG. In an advanced version, the game characters would themselves be sentient, for that extra dose of realism! Are you a game character, or a player character? :-)

By the way, the author of the article John Tierney gives the simulation idea a probability P of greater than 50%, while Bostrom, the Oxford philosopher who apparently thinks about this stuff as his day job, gives it about 20%. To me it's not implausible, but keep in mind: if you are *inside* the simulation your local conditions in principle tell you nothing about the outside world in which the simulation runs. So, we basically know nothing about P unless (1) we just happen to be in a realistic historical simulation or (2) we haven't yet hit the singularity. (2) is of course highly unlikely if R does go to infinity -- we'd be in a very special subset of sentient beings. So I'd say it's unlikely that our estimate of P is very good.

NYTimes: ...Dr. Bostrom assumes that technological advances could produce a computer with more processing power than all the brains in the world, and that advanced humans, or “posthumans,” could run “ancestor simulations” of their evolutionary history by creating virtual worlds inhabited by virtual people with fully developed virtual nervous systems.

Some computer experts have projected, based on trends in processing power, that we will have such a computer by the middle of this century, but it doesn’t matter for Dr. Bostrom’s argument whether it takes 50 years or 5 million years. If civilization survived long enough to reach that stage, and if the posthumans were to run lots of simulations for research purposes or entertainment, then the number of virtual ancestors they created would be vastly greater than the number of real ancestors.

There would be no way for any of these ancestors to know for sure whether they were virtual or real, because the sights and feelings they’d experience would be indistinguishable. But since there would be so many more virtual ancestors, any individual could figure that the odds made it nearly certain that he or she was living in a virtual world.

The math and the logic are inexorable once you assume that lots of simulations are being run. But there are a couple of alternative hypotheses, as Dr. Bostrom points out. One is that civilization never attains the technology to run simulations (perhaps because it self-destructs before reaching that stage). The other hypothesis is that posthumans decide not to run the simulations.

“This kind of posthuman might have other ways of having fun, like stimulating their pleasure centers directly,” Dr. Bostrom says. “Maybe they wouldn’t need to do simulations for scientific reasons because they’d have better methodologies for understanding their past. It’s quite possible they would have moral prohibitions against simulating people, although the fact that something is immoral doesn’t mean it won’t happen.”

Dr. Bostrom doesn’t pretend to know which of these hypotheses is more likely, but he thinks none of them can be ruled out. “My gut feeling, and it’s nothing more than that,” he says, “is that there’s a 20 percent chance we’re living in a computer simulation.”

My gut feeling is that the odds are better than 20 percent, maybe better than even. I think it’s highly likely that civilization could endure to produce those supercomputers. And if owners of the computers were anything like the millions of people immersed in virtual worlds like Second Life, SimCity and World of Warcraft, they’d be running simulations just to get a chance to control history — or maybe give themselves virtual roles as Cleopatra or Napoleon. ...

Strings in the New Yorker

I think the article, which discusses the new books by Woit and Smolin, is very fair, and it ends with a surprisingly mature recapitulation of the decoupling theorem and the irrelevance of quantum gravity to applied science. Every man, woman and child should read it and then ask their local particle theorist for more clarification.

I've read chunks of Smolin's book and it's quite good, although not without flaws. I have to go back and reread it -- I picked it up in a bookstore and couldn't put it down for at least an hour. I found his discussion of finiteness of string perturbation theory confusing -- he represents Mandelstam as saying one thing in the main text, but the email quoted in the footnotes doesn't seem to back it up. If the analytic continuation is the only problem then it's not on worse footing than many other results in theoretical physics. But, then, what have D'Hoker and Phong (and my grad school colleague Nathan Berkovits) been up to all this time? If I believed in string theory I'd have to spend some time sorting this all out.

...Today, more than a decade after the second revolution, the theory formerly known as strings remains a seductive conjecture rather than an actual set of equations, and the non-uniqueness problem has grown to ridiculous proportions. At the latest count, the number of string theories is estimated to be something like one followed by five hundred zeros. “Why not just take this situation as a reductio ad absurdum?” Smolin asks. But some string theorists are unabashed: each member of this vast ensemble of alternative theories, they observe, describes a different possible universe, one with its own “local weather” and history. What if all these possible universes actually exist? Perhaps every one of them bubbled into being just as our universe did. (Physicists who believe in such a “multiverse” sometimes picture it as a cosmic champagne glass frothing with universe-bubbles.) Most of these universes will not be biofriendly, but a few will have precisely the right conditions for the emergence of intelligent life-forms like us. The fact that our universe appears to be fine-tuned to engender life is not a matter of luck. Rather, it is a consequence of the “anthropic principle”: if our universe weren’t the way it is, we wouldn’t be here to observe it. Partisans of the anthropic principle say that it can be used to weed out all the versions of string theory that are incompatible with our existence, and so rescue string theory from the problem of non-uniqueness.

...Neither Smolin nor Woit calls for the forcible suppression of string theory. They simply ask for a little more diversity. “We are talking about perhaps two dozen theorists,” Smolin says. This is an exceedingly modest request, for theoretical physics is the cheapest of endeavors. Its practitioners require no expensive equipment. All they need is legal pads and pencils and blackboards and chalk to ply their trade, plus room and board and health insurance and a place to park their bikes. Intellectually daunting as the crisis in physics may be, its practical solution would seem to demand little more than the annual interest on the rounding error of a Google founder’s fortune.

“How strange it would be if the final theory were to be discovered in our own lifetimes!” Steven Weinberg wrote some years ago, adding that such a discovery would mark the sharpest discontinuity in intellectual history since the beginning of modern science, in the seventeenth century. Of course, it is possible that a final theory will never be found, that neither string theory nor any of the alternatives mentioned by Smolin and Woit will come to anything. Perhaps the most fundamental truth about nature is simply beyond the human intellect, the way that quantum mechanics is beyond the intellect of a dog. Or perhaps, as Karl Popper believed, there will prove to be no end to the succession of deeper and deeper theories. And, even if a final theory is found, it will leave the questions about nature that most concern us—how the brain gives rise to consciousness, how we are constituted by our genes—untouched. Theoretical physics will be finished, but the rest of science will hardly notice.

String Landscape, AI and virtual worlds

The Landscape lectures at Erice are over. I'm not very optimistic about the anthropic principle --- for example, see my paper with Graesser, Jenkins and Wise on how even Weinberg's original anthropic prediction of the cosmological constant is critically sensitive to the assumed level of initial density perturbations. Nevertheless I learned something interesting from Lenny's lectures. Since the Landscape involves many metastable vacua, it is very likely that our universe, with its small vacuum energy, was produced via bubble nucleation from a parent universe with much larger vacuum energy. In this case it is unavoidable that the spatial geometry of our universe has negative curvature. A subsequent era of inflation can reduce this negative curvature to almost zero, but there is a definite prediction that k is negative. If an observation by Planck (next generation CMB probe after WMAP) yields a small positive k, the Landscape will be strongly disfavored. By the usual Popperian definition, the Landscape is then science: it makes a falsifiable prediction! (Lenny, being quite honest, confided that Andre Linde would almost certainly find a way out, but it would have to be quite contrived :-)

[See more recent post here. The prediction of negative curvature isn't as robust as I had originally thought. 10.24.2006]

Having argued that the Landscape is sort of falsifiable, let me now turn to a reductio ad absurdum of anthropic arguments.

Let R = the ratio of number of artificially intelligent virtual beings to the number of "biological" beings (humans). The virtual beings are likely to occupy the increasingly complex virtual worlds created in computer games, like Grand Theft Auto or World of Warcraft (WOW will earn revenues of a billion dollars this year and has millions of players). In the figure below I have plotted the likely behavior of R with time. Currently R is zero, but it seems plausible that it will eventually soar to infinity. (See previous posts on the Singularity.)



If the integral of R diverges, then anthropic reasoning suggests that we are overwhelmingly likely to be virtual beings living in a virtual world. The electron mass and string flux compactification parameters were set by a programmer (himself virtual) working for a game company within yet another simulated world :-)