Ecology, Economics, and Exponentials: Modeling Technological Goals and Human Nature
[Editor’s Note: George Church’s presentation made extensive use of charts and other visual representations. A verbal summary is not able to convey his general ideas or his specific points adequately. The following summary, therefore, will necessarily appear disjointed and even incoherent in places.]
My talk will involve the convergence of genomics and computer science, with an emphasis on what sorts of timelines are plausible. I will also discuss the economic consequences, not only in terms of dollars, but also environments, ecologies, and so forth. I would like to give you a feeling for the kinds of things we do in systems biology, and then go on to discuss speculatively some implications it has for human nature. My background is in modeling, not only of the evolution of biopolymers such as proteins and DNA, but also structures that constitute molecular machines, and eventually whole ecosystems.
I have lived through two small revolutions, not as momentous as the Galilean and Darwinian, but quite important nonetheless: recombinant DNA and genomics. Both are reductionistic, and both rejuvenated systems biology, an old discipline. We now think of molecular processes as machines, but we can integrate them into metazoans like ourselves, cancerous stem cells, and ecosystems.
One way of modeling is to plot some calculated property on one axis and another observed property on another. Then we look for outliers, which are not to be swept away. They are not indicators that our model has failed. They are our friends and are potential discoveries about how our methods are not working. Some of these correlations have very interesting deviations from optimality. We are interested in how to take this ten-thousand-year history of genetically manipulating single molecules and, when we have a satisfactory level of precision, get to diversity. Despite doubts about what is acceptable in human breeding terms, it is quite possible that there is a great tolerance for diversity.
We think about diversity especially when we consider the limits to what we are able to do, and then try to extend them. Running speed, depths to which people can dive, breadth of the visible wavelength, temperature people can endure, the length of memory are all examples of limits that interest us. Once we exceed these limits, are we still within the range of human nature as we know it? We should remember that the Darwinian breakthrough, or clusters of Darwinian breakthroughs, allowed us to become hyper-adaptable and no longer dependent on DNA for inheritance and evolution. We are no longer limited by our germ-line. For some people laptops are as much a part of their being as their DNA. Maybe we should not change their germ-line but their laptops.
Does germ-line engineering hold the fastest promise for change? I would argue that somatic engineering is much closer at hand. Change through germ-line engineering takes about twenty years to manifest the result. On the other hand, somatic engineering—putting inorganic prosthesis or organic chemicals or somatic cell genetics into or outside our bodies—can take mere days. The other problem with germ-line engineering is the ethics of allowing adults to choose for their children or for other adults. Using the phenotype is more predictable. You can choose among a series of fertilized eggs, but that’s not very predictable. By the time to get to adulthood, on the other hand, it’s very clear what prostheses and drugs will and will not work. Similarly, if we want to work on our cells, histo-compatible adult stem cells may be more accessible and more appropriate, as will interfaces with organic engineering, nanotechnology, and more ordinary inorganic engineering.
When we plot growth in order to predict timelines, we see that the number of CPUs or CPU power will certainly overtake that of the population. CPU growth is definitely steeper than exponential growth and closer to a parabolic fit. When we compare the processing power of computers and the human brain, using Moore’s Law, we see a cross-over point at about ten to the fourteenth instructions per second. This does not address when or whether the entire internet will be equivalent to human intelligence. In addition to physical limits, we also need to think about cost limits and to compare them with other programs and their benefits, like launching satellites, or eradicating disease or sequencing the human genome. While undertaking any project always costs something, we should also remember than not undertaking them also entails a cost, which can often be considerable. Five percent of the global gross domestic product is dedicated to hackers and e-viruses.
The cost of progress, however, is not measured only in dollars. There is also the question of its effects on the environment. Since humans are hyper-adaptable, the amount of computing that we and our machines will be able to do is going to be limited primarily by the amount of energy at our disposal. There is, of course, sunlight and other sources of energy, like nuclear power. But we still have the problem of getting the heat off the earth. Right now we are consuming within three or four logs of the maximum. But not all energy is used for computation, and the efficiency of computation can change. In addition, there are alternative natural mechanisms that are doing similar jobs that are just now being discovered.
What will happen to diversity? There is the problem that a replicating system will turn the entire surface of the earth into itself. But there are other, more insidious ways of losing diversity. We may have to start thinking seriously about using geographical isolation to achieve this goal. It has played one of the major roles in evolution. We need to know not only common mutations, but every mutation that occurs, not just in our germ-line, but in our somatic cells as well. One problem here is devising the instrumentation that can monitor these changes and making it inexpensive enough to use.
We have seen maybe four logs of improvement in efficiency over the course of the genome project, but we’re still ten logs away from the efficiency of some very commonly used equipment, like video recording. We would like to be able to get DNA analysis and nucleic acid analysis down to the level of video recording costs. We have discovered that existing organisms can do inorganic and organic nanofabrication beautifully. We need to harvest the biosphere for these remarkable molecular machines. In a certain sense, therefore, we already have achieved atomic precision.
I’ve been talking about systems biology because it is so embracing and more holistic than our usual speculations. The timelines I’ve been discussing are often higher than exponential. They may not continue at these rates, but their limits are determined by mass and cost. We should start thinking about our inheritance in larger terms than DNA. Germ-line changes are the least of our worries.
Biosphere Technologies and the Myth of Individuality
The “Gaia Hypothesis” explains the tendency of the Earth’s surface to maintain its temperature, reactive gas concentration, and alkalinity within astronomically narrow limits for millions of years. The self-maintaining properties of cells, organisms, communities, and ecosystems can be extrapolated to the atmosphere and surface sediments of planet Earth. Not only are we people (Homo sapiens mammals), one of the more than 10 million existing species components of the Gaian regulatory system but so are our machines. I argue that although not by themselves alive, like viruses and beehives, machines are capable of growth, reproduction, mutation, and therefore evolution. Machines change through time. Even though they are not self-sustaining and they have no metabolism, machines do evolve.
No single species is privileged. Many populations of organisms, like us, disrupt their own habitats by outgrowing their own ecological support systems. The Gaian Earth-regulating system which responds to perturbation by changes in metabolism, differential survival, growth, and species origin and extinctions maintains dynamic stability of the planet’s surface. The fossil record informs us that, for members of any given species, habitat loss is followed by population decline and, eventually, by extinction.
Technohumans grow now as “mammalian weeds.” Non-human ecosystems are converted to the agro-urban-technological, primarily by water and solar-radiation rerouting, soil depletion, and fossil-fuel combustion. The extremely successful recent human reproductive strategies alter or even extinguish lacustrine, riparian, dunal, marine coastal, forest, grassland, chaparral, and other non-human, primarily terrestrial, ecosystems. The agro-urban network overgrowth adds cellulose, hemicellulose, lignin, polymeric plastics, metal oxides, aldehydes, aromatics, and hundreds of other compounds and sedimentary particles even to ocean water. The accelerated patterns of surface transformation to the agro-urban network coupled with Homo sapiens-induced species extinction are reminiscent of a phenomenon at a smaller size-scale: malignant melanoma and other solid tumor metastasis. Although ecological alteration of the Gaian body politic is about 105 times larger than melanoma or other cancers, the two phenomena share at least these characteristics: uncontrolled growth stimulated by the prototactic imperative to reproduce, and metabolic dependency on surrounding supportive communities by rerouting of energy, fluids, and organic compounds to the sites of most rapid proliferation. Lack of neural or other centralized control unleashes destructive, compulsive, proliferative behavior in cells, tissues, organisms, populations, communities, and beyond. Although we perceive ourselves, usually as individuals, scientific analyses shows each of us to be at least 10 percent dry weight “foreign” (i.e., bacterial). By means of a video (I hope) we will see how we, like all “individual animals” are complex composites, integrated communities that require chemical, microbiological, and ecological studies to be properly understood.
[Editor’s Note: A large part of Professor Margulis’s presentation consisted of slides and a long film of various microorganisms. Some of her commentary which is comprehensible only in conjunction with its illustrations has therefore regrettably been omitted.]
I feel so humble in the face of the past that I cannot talk about the future. The past, the evolutionary past, is so complex that thinking about the future in technological terms just boggles my mind. Emily Dickinson wrote:
A little Madness in the Spring
Is wholesome even for the King,
But God be with the Clown—
Who ponders this tremendous scene—
This whole Experiment of Green—
As if it were his own!
That’s how I feel about this meeting. It’s clowning to think that we can predict in detail the carbon dioxide of the atmosphere or the germ-line.
I want to impress you with the fact that technology belongs to the biosphere—what traditionally is called the noosphere. I want to show you some technologies that are extremely ancient. There is, for example, architectural habitat alteration that controls light, temperature, chemical composition, and water better than this room does. The one thing that distinguishes our species from the rest of nature is speech and symbolism. We can talk, and therefore we can lie. Deceit is all over biology.
When we look at these ancient natural technologies and then think about what I call the myth of individuality, we begin to appreciate the extent to which organisms are composite. We are also going to see how the World Wide Web, a communication among what looks like individual organisms, has been on the earth for maybe 35 hundred million years. It’s not silicon technology, but it is true technology. The idea that humans can synthetically adopt and incorporate photosynthesis into themselves is a real possibility and already exists in nature. Some predatory animals have developed associations with very efficient photo-synthesizers.
Examples of biospheric technology are the large termite mounds in Africa. Temperature there is regulated to within half a degree centigrade, humidity is maintained at 95 percent in extremely dry surroundings, and there is as much of the termite mound below ground as there is above it. Air flows through it, and there are divisions maintained into morgues, school rooms, and hatcheries. The termites derive their source of carbon and energy from the fungi they have learned to grow as crops. There can be as many as 30 million termites in these mounds, along with all sorts of associated animals that live with them.
We share 99.9 percent of our DNA with chimps. From a biological point of view, we are just another chimp. If we want to think about technological potentials for ourselves, we might take a look at how problems have already been solved in nature. Here is a stromatolite from the sea floor. It is the product of bacteria and has lasted hundreds of thousands if not hundreds of millions of years. It is a very complex community of microorganisms that are constantly maintaining and stabilizing sediment and recycling carbon and phosphorus in ways we have not even begun to approach. There are also examples of web organisms that communicate with each other, and photo-synthesizers that produce carbon and energy for the rest of the community and recycle the sulphur. These are stable, worldwide communities. We don’t know how they’re communicating, but they are communicating well enough for the same composition to be fundamentally the same worldwide.
As for the subject of individuality, here are mollusks that live entirely by photosynthesizing, having incorporated photosynthesis into their own bodies. Some animals incorporate the photosynthetic chloroplasts of algae. Others actually focus light on the photosynthetic entities that support it. These mechanisms suggest more feasible technologies than changing the germ-line of people.
[Professor Margulis showed a video and commented on it.]
Any organism has a multiple genomic background. The theory of the origin of species does not really lie in mutation. Mutations just modify. If you want to change organisms in serious ways, you can think about acquiring and integrating genomes that have already been optimized by natural selection. I want Emily Dickinson to have the last word:
But nature is a stranger yet;
The ones that cite her most
Have never passed her haunted house,
Nor simplified her ghost.
To pity those that know her not
Is helped by the regret
That those who know her, know her less
The nearer her they get.
This is a chapter from The Future of Human Nature: A Symposium on the Promises and Challenges of the Revolutions in Genomics and Computer Science (Conference).
Previous: Session Four | Table of Contents | Next: Session Six