Chapter 113 19.2 Natural selection is not enough

Naturalists around the world make long-term observational studies of the evolution of populations of life in wild environments: snails in Tahiti, fruit flies in Hawaii, songbirds in the Galapagos Islands, and lakes in Africa. fish.As the research progresses year by year, scientists have a better chance of clearly demonstrating that long-term evolution has been going on in the wild.Shorter-term studies using bacteria and, more recently, the face weevil have shown in the laboratory what short-term evolution of organisms looks like.So far, the results of these experiments with living populations of organisms have been consistent with what neo-Darwinian theory predicts.The beaks of songbirds on the Galapagos Islands did thicken and thicken over time in response to drought-induced changes in food supply, as Darwin predicted.

These careful measurements confirm that self-managed adaptations do occur spontaneously in nature.They also clearly demonstrate that when the trivial but steady removal of "not applicable" parts increases, their accumulation can naturally show significant changes.However, the experimental results did not show new levels of diversity or any new species, nor did they even prove that new complexity emerged. Despite our careful scrutiny of the historical record, there is no record of a new species evolving in the wild.Moreover, the most notable thing is that in the process of human domestication of animals, no new species have appeared.This includes that no new species of fruit flies have appeared in the study of hundreds of millions of generations of fruit flies, and people have deliberately added environmental pressure to the fruit fly population in order to induce the formation of new fruit fly species.In the realm of computer-simulated life, the word "species" has little meaning—except for initial bursts, there have been no cascades of entirely new species.In wild environments, captive breeding environments, and artificial life environments, we have all seen the emergence of mutations.But, in the absence of greater changes, we are also well aware that the range in which variation occurs appears to be narrow and often confined to the same species.

Regarding this phenomenon, the standard explanation is that we are actually measuring an event that occurred in a long geological time with a ridiculously short time span. So, what else can we expect to see?Life existed in a bacteria-like form for billions of years before undergoing cataclysmic changes.Please be patient!That's why Darwin and other biologists turned to the fossil record to provide proof of evolution.But while the fossil record indisputably demonstrates Darwin's more important claim—that changes in traits accumulate over time in offspring—it fails to show that these changes are attributable purely to natural selection, or even that the changes should be primarily Credit to natural selection.

Because, so far, no one has witnessed the exact moment of change in the fossil record or in real life, or in computer simulations of artificial life—the moment when the mechanisms of natural selection spur its complexity to a new level.There seems to be some sort of dubious barrier between neighboring species that either prevents this crucial change from happening or moves it out of our view. Steven Jay Gould argues that the incredible instantaneousness of evolution (in the context of evolution) removes exact stages of change from what we perceive as the fossil record.Whether his theory is correct or not, the available evidence suggests that there is some natural constraint that prevents the propagation of small changes, and that evolution must try to overcome this constraint.

Artificially synthesized protolife and computer-simulated artificial evolution have brought us more and more surprises.However, artificial life suffers from the same drawbacks as its cousin artificial intelligence.As far as I know, no artificial intelligence—be it an automaton, a learning machine, or a large cognitive program—runs continuously for more than 24 hours.After a day and a night, these artificial intelligences will stop functioning.The same goes for artificial life.Most of the simulated lives that rely on calculations to run, after a while of excitement, soon fell silent.Although programs are still running at times, churning out minor changes, they do not leap to new levels of complexity or generate new surprises after the first climax (including Tom Ray's "Earth").Maybe give them some runtime and they'll work.But, for whatever reason, computer-simulated lives generated by naive natural selection do not experience the magic of free evolution, which their creators and I would love to see but not see.

As the French evolutionist Pierre Grass said: "Variation is one thing, evolution is quite another. The difference between the two cannot be overemphasized... Mutation provides change, but Not progress.” So, while natural selection may have produced microchanges (some trending variations), there is no guarantee that it can produce macrochanges, the freedom to create unforeseen new shapes and processes toward growing complexity . Even if artificial evolution is only a slight change in adaptation, the artificial evolutionary prospects predicted in this book will still be realized.Self-directed mutation and selection mechanisms are powerful geniuses for tackling difficult problems.Over small spans of time, natural selection does work.We can use it to find evidence we can't see, to fill gaps we can't imagine.The question boils down to whether random mutation and selection mechanisms alone are sufficient to produce new things consistently over long periods of time.Also, if it is true that "natural selection alone is not enough", what other forces are at play in natural evolution?What else can we introduce into artificial evolution to produce self-organizing complexity?

Most critics of natural selection grudgingly admit that Darwin's "survival of the fittest" was correct.Natural selection means primarily the destruction of the unfit.Once a suitable person is produced, the momentum of natural selection to eliminate the inferior will be unstoppable. But creating something useful sounds a little daunting.What the Darwinian view ignores are plausible explanations for fit.Where is the fit before being selected?According to the prevailing neo-Darwinian interpretation today, the fit is due to random variation.Random variation within chromosomes causes random variation in the developing organism, which in turn adds fitness to the organism as a whole from time to time.In other words, fits are randomly generated.

This simple process can induce coordinated change in a relatively short period of time, as demonstrated by field as well as artificial evolution experiments.But if natural selection were able to weed out an infinite number of failed attempts and for an infinite amount of time, could this random variation produce a complete series of its desired winners to choose from?Darwinian theory is tasked with proving that the combination of the negative braking force of death selection and the random, aimless forces of chaos can produce a sustained, creative, positive push towards the great The enduring complexity of nature over eons.

Post-Darwinism proposes that, after all, there are other forces at play in the evolutionary process.These authoritative mechanisms of change reorganized life into new fitness.These unseen agents expand the repository of life's information, and perhaps that is the repository of natural selection.Deep evolution is not necessarily any more mysterious than natural selection.They see each dynamical symbiosis, directed mutation, jump theory, or self-organization theory as a mechanism, a mechanism that, in the long run, complements Darwin's relentless selection process to drive evolutionary innovation. mechanism.