Chapter 6 Chapter Three Exploring the Devil's Secret

When Kaufman learned all this, he was in awe.Order appeared again, a naturally formed order!It is so beautiful that words cannot describe it.But is this life? In the fall of 1986, while Anderson and Arrow were drafting the invitation list for the economic conference, Cowan signed a three-year lease with the Diocese of Santa Fe for the Christopher Monastery on Winding Canyon Road. (Christo Rey Convent).It's a mud-brick bungalow next to an expensive lot lined with galleries. It's almost time.Thanks to the operating funds allocated from institutions such as the MacArthur Foundation, Ke Wen and his colleagues have begun to hire several staff for the institute, and the staff desperately needs their own offices.More importantly, the economic conference will be held soon, and several other seminars are also being planned. The institute urgently needs some office space, so that when scholars come to visit, they can have a desk and a phone to use.Ke Wen felt that although the monastery was small, it was still usable, and the rent was so cheap that people were reluctant to give it up.So, in February 1987, the Institute officially moved, and within a few days, this small space was filled.

chaos The crowded situation never improved.On Monday, August 24, 1987, the first time Arthur stepped through the gates of the Santa Fe Institute, he almost stumbled onto the receptionist's desk, which was crowded by the entrance. , barely an inch from the open door.The hallway is full of books and paper boxes, the photocopier is stuffed in the cabinet, and a staff member is not working in the hallway at all.The place was chaos, but Arthur fell in love at first sight. There is no better place for me than here!He said.In peace, secrecy, and serenity, this sprawling monastery also manifests an intellectual vitality.Richardson, the event director, came out to greet him and show him around, stepping on the wrinkled, canvas floors, admiring the lovely craftsmanship on the doors, the gleaming lampshades and the intricate ceiling decorations.She told Arthur how to get to the Eisenhower kitchen to get coffee: You had to go through the abbot's office, where Cowan now works.What used to be the church is now the meeting hall; on the other end of the wall, where the altar was once, now hangs a blackboard covered with equations and diagrams, reflected in the flickering light of the stained glass.She then showed him the cramped visiting scholar's office, which had been converted from the original nun's dormitory, crammed with cheap metal desks and typist's chairs, and looked out on to a courtyard bathed in sunlight. You can also overlook the Sangre de Cristo Mountains.

This was Arthur's first time in New Mexico, and he quickly became captivated by the beauty.The shock of the surrounding distant mountains, the bright sun of the desert, and the crystal-clear desert scenery is no less than that of painters and photographers when they first saw this scene.However, he felt at once that there was a special magic in this monastery.The whole atmosphere was unbelievable, Arthur said: When I saw the various books on display, the various papers scattered here and there, the free, unrestrained atmosphere, I couldn't believe that there were places like this in the world.He began to think that this economic seminar would be very stimulating.

hit it off The visiting scholars were stuffed into the office in groups of three or two, and handwritten pieces of paper were pasted at the door as name tags.One office door had a name Arthur was interested in: Stuart Kauffman from the University of Pennsylvania.Arthur had a quick meeting with Kaufman two years ago at a conference in Brussels, where he was impressed by Kaufman's talk of growing embryonic cells.His thesis was that cells send out chemical messages that drive the development of cells in other embryos, thus producing a unified organism rather than just a mass of protoplasm.This coincides with Arthur's idea of ​​human society as a self-unifying, mutually supportive, interactive relationship.He still remembers coming home and telling his wife, Susan: I just had the best speech I've ever heard in my life!

So, as soon as he was settled, he wandered over to Kaufman's office.He said: Hi, do you remember we met two years ago? Well, don't remember, in fact Kaufman forgot.But come in!With his bronzed skin, curly hair and Californian ease, Kaufman, 48, is very approachable.Arthur, too, was in such a good mood that morning that he would have been glad to see anyone.The two hit it off.Arthur said: Kaufman is a very warm person, you want to hug him when you see him, and I usually don't want to hug people when I meet him.But he is just so likable! Of course, they quickly started discussing economics.The meeting is about to be held, and they are naturally very concerned about this topic, but they can't predict the possible situation in the meeting at all.Arthur began to tell Kaufman a little about his research on increasing returns.Arthur laughed: This gave Kaufman a great opportunity to corner me and tell me his latest thinking.

Arthur soon discovered that Kaufman was a creative person, like a composer with all kinds of melodies in his head.He came up with new ideas in an endless stream, and he talked far more than he listened.Speaking his thoughts out loud seemed to be a way of thinking for him.And then he'll go on and on about his thoughts, about his thoughts. breathtaking scenery The previous year, Kaufman had been haunting Santa Fe, and people at the institute had long been familiar with his behavior.Kaufman's father was a Romanian immigrant who amassed a small fortune in real estate and insurance, so he was one of the few scientists who could afford to set up another home in Santa Fe, where he lived for the first half of each year.In every preparatory meeting of the institute, Kaufman would, in his confident baritone voice, spout out a bunch of suggestions.During the question and answer time of each seminar, he will also hear him aloud to make suggestions on that topic: we imagine a group of light bulbs connected together randomly, and then during the break period in the middle of the meeting, he will even hear him looking for people to talk about his new ideas .Legend has it that he was heard lauding some of his insights in theoretical biology to workers repairing photocopiers.If there is no visitor, he will quickly grab a colleague who is closest to him and repeat in detail what he has said a hundred times.

All this was enough to drive his best friend away, yelling that he couldn't take it anymore.To make matters worse, Kaufman thus became known as overly cocky, garrulous and insecure, though some colleagues were quick to add that they liked Kaufman anyway.Regardless of how people feel about Kaufman, Kaufman is Kaufman.For twenty-five years, he has been firmly grasping a beautiful scenery, which is so shocking and beautiful that he can't help himself. Order can only be used to describe his ideas, but order is not enough to capture his true thoughts.Listening to Kaufmann talk about order is like hearing someone express primitive mysticism in the language of mathematics, logic, and science.For Kaufman, order can unravel the mysteries of human existence, explaining why living, thinking beings like us are possible in a universe that seems to be governed by accident, chaos, and blind laws of nature.For Kaufmann, order tells us that humans may indeed be an accident of nature, but at the same time they are more than an accident.

Kaufman is always quick to add that Darwin was absolutely right: humans and other living things are the result of four billion years of random mutation, catastrophe, and struggle for survival, and humans are not miracles or space aliens.But he emphasized that Darwin's theory of evolution cannot be the whole picture. Darwin did not understand the principle of self-organization. Matter will continue to self-organize into more complex structures, even in the face of the endless destructive force in the second law of thermodynamics. in this way.Nor did Darwin know that the forces of order and self-organization are to the creation of living systems as they are to the formation of snowflakes or convection in simmering soup.So, in fact, the story of life is a pile of accidents and contingencies, Kaufman said, but it is also a story of order: a tapestry of nature woven by inner, deep creativity.

I just love the story, says Kaufman: I've spent my life trying to figure out how to get this story right. mystery of order Walk down the corridors of any scientific institution in the world, and you hardly go far to see an office with a poster of Einstein: Einstein wrapped in a coat, walking absently through the snow in Princeton; Einstein Staring at the camera emotionally, with a pen tucked in the collar of his worn-out sweater; Einstein smiled narrowly and stuck out his tongue at the world.The creator of the theory of relativity is almost a common scientific hero all over the world. He symbolizes deep thinking and free creative spirit.

As early as the 1950s, Einstein was the hero of the young Kaufmann.I admire Einstein very much, and he said: No, I shouldn't use the word admiration.I should say love, I like that he regards theory as the free creation of the human mind, and science as exploring the secrets of the devil.Einstein used the devil as a metaphor for the creator of the universe.Kaufman particularly remembers his first exposure to Einstein's ideas in 1954, when he was fifteen years old, when he read about Einstein and Leopold Infeld in a bestseller. The origin of the commonly proposed theory of relativity.I was excited to understand for myself.Einstein's inventive genius and free mind allowed him to create a world in his head, and I thought it was amazing that someone could do that.I still remember when Einstein died in 1955, I cried, as if I had lost an old friend.

Before reading that book, Kaufman had been a good A or B student.From then on, his passion was ignited, but not necessarily the enthusiasm for science. He didn't feel that he had to follow Einstein's footsteps step by step, but he did feel the same strong desire to get a glimpse of the deep mysteries of the universe .When you look at the work of a cubist artist and see the hidden structure inside, that's what I wanted to explore.His immediate interest was not at all in science, and the young Kaufman aspired to be a playwright, exploring the light and dark sides of the human soul.His first attempt was an opera with Todd, his high school English teacher, and it was botched!However, for Kaufman, an adult (the English teacher was twenty-four years old at the time) took his inner excitement seriously, which was a crucial step in enlightening his awakening of knowledge.Although this is not a very good opera libretto, if I can write an opera with my teacher at the age of sixteen, what else can I not do? Therefore, Kaufman, who entered Dartmouth as a freshman in 1957, was a playwright in every cell.He even smoked a pipe because friends told him he had to smoke one if he wanted to be a playwright.Of course, he continued to write plays, and that year, he wrote three plays with his classmate and friend McGory. However, Kaufman soon noticed a peculiarity of his play: the characters were often preaching.They talk about the meaning of life and what it means to be a good person, but they just sit around and don't act.He began to understand that he was far more interested in the ideas of the characters than in the script itself.Even though I don't know exactly what I want, I know I want to find my own way into that powerful and wonderful hidden world.When I found out later that my friend Greene, who was at Harvard, was going to major in philosophy, I was devastated.I wish I could be a philosopher too, but of course I have to be a playwright; and to give up the theater is like giving up the identity I had begun to make for myself. From Philosophy to Science After struggling for a week, he suddenly realized: I don't have to be a playwright, I can be a philosopher!So, for the next six years, I devoted my great enthusiasm to studying philosophy.Of course, he started with ethics, and when he was reading plays, he always wanted to understand the question of good and evil.However, he soon moved on and became interested in the philosophy of science and the philosophy of mind.To me, this seems to be where the deep mystery lies.He said.Why can science discover the nature of the world?And why can the mind understand the whole world? With such enthusiasm, Kaufman graduated third in 1961 and received a Mashall Scholarship from Oxford University to continue his studies.As a result, he took a detour.I had eight months to spare before I could report to Oxford, so I did the only sensible thing: I bought a car, lived in the Alps and skied.I have the address of the most prestigious Post Hotel in St. Anton, Austria, I park my car in the hotel parking lot and use their restroom quite often throughout the winter. Upon arriving in Oxford, he had an immediate appreciation of his surroundings.He recalled three times in his life when he had been in the mind-boggling temples of knowledge, Oxford being his first experience.For the first time in my life, I was surrounded by people smarter than me, and there were a lot of talented Americans who went there, some of whom are still famous, like David Souter, who used to be with us and is now on the Supreme Court.And Will (George F. Will, a well-known American news critic and columnist) and I used to go to Indian restaurants to have tooth festivals together to avoid school meals. Kaufman was eager to understand science and the mind, so he took courses at Oxford called Philosophy, Psychology and Physiology.This course not only covers traditional philosophy, but also discusses the neural structure of the modern visual system and the neural wiring model of the brain.Simply put, this class discusses the workings of the mind from a scientific perspective.His psychological tutor named Su Delan (Stuart Sutherland) had a great influence on him.Sutherland likes to sit behind a desk and drill students with a barrage of mental gymnastics: Kaufman!How does the visual system distinguish between two points of light projected on adjacent cones in the retina?Kaufman found he liked the challenge, and he had a talent for simulating answers on the spot, coming up with one that at least seemed plausible.In fact, he admits that impromptu simulations have become his habit ever since. Ironically, it was this talent for simulation that led him to abandon philosophy in favor of the more practical academic medicine. would rather be einstein The way I make decisions proves that I can never be a great philosopher, he said with a smile: My reasoning is that I can never be as smart as Kant, and unless you are as smart as Kant, there is nothing wrong with being a philosopher Meaning, therefore, I should go to medical school.You see, this is not a syllogism at all. The real reason, he said, was that he had grown impatient with philosophy.It's not that I don't love philosophy, it's that I don't trust the frivolity of certain parts of philosophy.Modern philosophers, or at least philosophers in the 1950s and 1960s, were so obsessed with examining concepts and what they implied, that they neglected the truth.So, you can test whether your argument is pertinent, appropriate, consistent, etc., but you have no way of knowing whether you are right or not.In the end, I'm very unhappy with this. He hopes to unearth the truth and learn the devil's secrets.If I had to choose, I would rather be Einstein than Wittgenstein (Ludwig Wittgenstein, 1889︱1951, a famous American philosopher). More importantly, he mistrusts the frivolous side of his personality.I've always had an ability to conceptualize things.On the plus side, it's the deepest part of me, a great God-given gift, but on the minus side, it's just glib and superficial.Because of this consideration, I said to myself, I'm going to study medicine, and they won't make me slick and ostentatious.Because I have to take care of patients, they will force me to learn a lot of real things. True, but the med schooling didn't change Kaufman's penchant for entertaining himself with new ideas.Because he had never taken a medical preparatory course, in 1963, before he entered the University of California, San Francisco, he applied to Berkeley for a year of courses.It was also in Berkeley that he took embryology for the first time. He was taken aback.This phenomenon is too shocking.It starts with a tiny fertilized egg, and the thing slowly develops into orderly newborns and adults.Somehow, a single egg cell differentiates into nerve cells, muscle cells, and liver cells, among hundreds of other different cells, with incredible precision.The strange thing is not the occasional abnormal fetus, though that is a tragedy; the strange thing is that most babies are born perfect and sound.This remains one of the most beautiful mysteries in biology, and I was so utterly fascinated by cell differentiation that I decided to take a good look at it.He said. A computer that executes genetic instructions He is in his prime.Jacob and Monard just published a series of papers on genetic circuits between 1961 and 1963, for which they later won the Nobel Prize (Arthur was not published until 16 years later. read about it on a beach in Hawaii).Therefore, Kaufman quickly came into contact with their theory that every cell has many regulator genes, which can turn other genes on or off like switches.This discovery was an apocalypse for all biologists.If genes could switch on and off each other, then there would be genetic circuits.The genome is a kind of biochemical computer. The computational behavior of the whole system, that is, the orderly activities, determine the differences between cells. The question is: how? In fact, many scholars did not worry much about this problem until then (or until now).They discussed the developmental program in the cell, as if the DNA computer executes the genetic instructions step by step, just as the IBM mainframe executes the program written in FORTRAZ (a computer language) step by step.They also seem to believe that these genetic instructions are well-structured and have been corrected for errors by the process of natural selection, just like a human-designed computer program.Certainly so, the slightest error in the genetic program can cause a developing cell to transform into a cancer cell, or simply kill the cell's life.That's why hundreds of geneticists are working in the lab to decipher the biochemical mechanism of how gene A turns on gene B, and to study how the activity of genes C, D, and E affect this switching process.They believe that all the answers lie in these details. The more Kaufman pondered the image, the more question marks grew in his head.Yes, the genome is a computer, but it's nothing like an IBM-built machine.In real cells, many regulatory genes act simultaneously.Therefore, it is not as step-by-step as the computer made by humans. The genome computer must process most of the genetic instructions in parallel at the same time.Kaufman reasoned that if this is the case, what matters is not whether one regulatory gene stimulates the activity of another regulatory gene in strict accordance with the established order, but whether the entire genome can settle down and form a stable, self-unified form.The number of cycles in which a regulatory gene might go through at most two, three, or four different states is not that large, or the cell would be thrown into chaos as genes randomly switch on and off each other.Of course the genes active in liver cells would be very different from those active in muscle cells or brain cells, but maybe that's the point, Kaufman thought.A single genome can have many different active stable forms, which is perhaps why it produces many different cell morphologies during development. Unraveling the Mystery of Genetic Circuits The common unspoken assumption is that the details matter most.This also troubled Kaufman quite a bit.He knew that the details of biomolecules obviously mattered, but if genomes had to be perfectly organized and regulated to function, how could they have been created by evolutionary random trials?It's like honestly shuffling the cards and getting a hand of thirteen spades is not impossible, but unlikely.He said: It just doesn't feel right, how can you ask God or the process of natural selection to do this?If we had to explain the biological order in terms of improbable bits and pieces of the process of natural selection, if every object we looked up had struggled from the start, then we couldn't exist! He thought it should be more than that.I also hope that order has been established from the beginning, and there is no need to re-establish it or evolve it.I deliberately hope that the order in the regulatory system of inheritance is natural and natural.If this is the case, he reasoned, then this spontaneous, self-organizing quality of life runs counter to the theory of natural selection.According to Darwin, the precise genetic details of any organism are the product of random mutation and natural selection.But the organizational order of life itself is deeper and more fundamental.Life emerges purely from network structure, not genetic details.In fact, order is one of the devil's secrets. I don't know where the impulse came from, why am I just so curious about this question?It's a fantastic puzzle and I find it amazing that it breathes new life into my mind to be able to ask a question like this.But, I've kept that feeling all my life, and all of my dearly loved scientific research is about understanding this mystery. Indeed, for Kaufman, a twenty-four-year-old pre-med student, order is like a lingering itch.He wondered what it meant for the genetic order to exist freely?Take a look at genetic circuits in real cells!They've obviously been honed over millions of years of evolution, but beyond that, are they really unique?Of the innumerable possible genetic circuits, are they the only ones capable of producing orderly and stable states?If that's the case, then it's like getting a poker hand full of spades.If only relying on evolution, can it really be so lucky to produce the only possibility?This can only be described as a miracle.Or, stable circuits are not as rare as all-color cards, but as common as mixed-color cards such as spades, clubs, hearts, and red bricks?Because if that's the case, then it's much easier to hit right by chance in the course of evolution, and the genetic circuits in cells are just combinations that just happen to pass the rigors of natural selection. The only way to find out is to shuffle the cards, that is, to experiment with many genetic circuits that form, to see if they actually lead to a stable form?I immediately thought, what would happen if thousands of genes were randomly connected? He knew how to think about it: he had read Neural Circuits at Oxford.Of course, real genes are complex.But Jacobs and Monard tell us that a regulatory gene is basically just a switch, meaning that it switches between active and inactive states.Kaufman likes to think of them as light bulbs (on|off) or logical statements (true|false).Whatever imagery is used, the behavior of this switch captures the essence of the gene, leaving only the network of interactions between genes to be studied.So, when Berkeley's freedom of speech movement was in full swing on campus, Kaufman spent his spare time on the top floor of the apartment building, obsessively drawing one after another genetic network, wanting to understand how they Switch each other on and off. Self-contained form on and off Kaufman was so obsessed with solving this mystery that he continued to obsess over it even as he packed up his pre-med program to study medicine in San Francisco.It wasn't that medical school bored him; on the contrary, he found it very, very difficult.The professor either asked them to memorize a lot of things by rote, or asked them to work endlessly to analyze the physiological structure of the kidney and so on.At the time, he still hadn't wavered in his ambition to practice medicine, which appealed to his inner Boy Scout: being able to help others and knowing how to handle any situation, like pitching a tent in a storm. Kaufman can't help but be addicted to online games.I am eager to study the strange science of this random network.As a result, he only got a C in pharmacology.My notebooks are filled with graphics of genetic networks. At first, he found these circuits very confusing.He knew abstract logic but next to nothing about mathematics, and the computer textbooks he found in the library were almost entirely useless.At that time, automata theory (automata theory) had been established, and this set of theories mainly talked about the switch network of logic.These books can tell me how systems are synthesized, or what are the general limits of complex automata; but I am interested in the natural laws of complex systems, where does order come from?At least as far as I know, no one thinks about it.So, he continued to draw a lot of network diagrams, hoping to feel the behavior patterns of the network from them.If mathematics was required, he tried to invent it himself. He soon discovered that if the network became entangled like a plate of spaghetti, with every gene dominated by many others, the entire system would be violently shaken.If you use the lightbulb analogy, it's like a giant Las Vegas-style sign whose wiring is out of order, and the lights on it are flickering randomly. Kaufman also believed that if each gene was dominated by at most one other gene, the network would be very loosely connected and its activities very pure.It's as if the light bulbs on the signs are just going on and off like dull flashlights.But that's not the order Kaufman had in mind. He wanted his genetic bulbs to organize themselves into interesting shapes, like the swaying of a palm tree or the graceful dance of an anthurium.Jacobs and Monard have shown that, in practice, each gene is usually dominated by several others. (Today, the known number is two to ten genes.) So Kaufman took the middle path, which is to study networks that are loosely connected, but not too loosely connected.In fact, for the sake of simplicity, he studied networks in which each gene had only two inputs—that is, it was governed by only two other genes.At this time, he began to discover some strange clues.He already knew that tightly connected networks can be hypersensitive, which means that if you change the state of one of the genes from on to off, it will pull out an avalanche of changes, endlessly churning through the network .This is why tightly connected networks are often chaotic and never stand still.But in his network of two inputs, Kaufman found that changing the state of one gene did not lead to an unstoppable wave of changes.Usually, the altered gene simply flips back to its original state.In fact, as long as the difference between the two forms is not too large, they will tend to be unified.Kaufman said: The situation gradually simplified.I could see the bulb tending to get stuck on or off.In other words, this two-input network is as if you let the light bulbs on the signboard blink at will, and they always organize themselves into patterns of flamingoes or champagne glasses. ask the computer to do it for you order!Using stolen time from class, Kaufman filled his notebooks with random networks of two inputs, analyzing the activity of each network in detail.The job is both fun and frustrating.The good news is that networks of two inputs almost always stabilize quickly, cycling through at most a few different states, which is the normal state of stable cells.The bad news is that he has no way of knowing whether this model has anything to do with the real genetic regulatory network, because the real network of cells contains tens of thousands of genes, and Kaufman's hand-drawn network map only needs to cover five, Six genes are already in a mess.To keep track of all possible states of a network of seven genes, a matrix of 218 rows by 14 columns would be filled.Analyzing a network of eight genes would again require a matrix twice as large; and so on.The opportunities for error in hand-drawn graphics are enormous, says Kaufman: I've been double-checking my seven-element network, and there's no way I can draw an eight-element network by hand. In my second year of medical school, I couldn't stand it, I had played enough.So I went to the computer center across the street and asked if they could write the program for me.They said: Sure, but you have to pay.So I pull out my wallet and I'm happy to pay. After deciding to hire a computer to do it for him, Kaufman decided to simulate a network of 100 genes.Looking back, Kaufman says with a smile, he had no idea what he was doing.A single gene has only two states: on and off.The network of two genes will have two times two, four states: on︱on, on︱off, off︱on, off︱off.A network of three genes would have two by two by two, eight states, and so on.So a network of one hundred genes represents a state multiplied by two by itself one hundred times, which equals almost a million trillion trillion, that is, a one followed by thirty zeros.That represents a myriad of possibilities, Kaufman said.In theory, his simulated network should roam randomly through this vast variety of possibilities. In other words, there is no hope of his cell cycle idea being proved; Only when the state changes, can there be a way to re-trace each step, which is simply beyond imagination.Kaufman said: If it takes a computer to move from one state to the next in a millionth of a second (microseconds), and you let the computer run for a million trillion trillion microseconds, it is equivalent to billions of times the history of the universe .There's no way I'm going to finish my medical degree!Indeed, the computer bill alone would have bankrupted him long ago. Fortunately, Kaufman didn't really need to perform such a huge calculation at the time.With the help of experts in the computer center, he coded the two input networks of the 100 genes to be simulated by the computer, and happily handed over a bunch of perforated data cards to the counter.Ten minutes later the answer came out, printed on large report paper.As he expected, the network quickly settled into an orderly pattern, with most genes resting on or off, while others cycled through several different forms.If his network of a hundred genes is like a Las Vegas bulletin board with a hundred light bulbs, then this state of order, like the pattern that appears on the board when you win the lottery, does exist, and very stable. Why is there order? I couldn't be more excited!Kaufman said: Until now I feel very mysterious, I found a phenomenon that no one has observed intuitively.His two-input network did not roam the vast expanse of a trillion trillion states, but quickly settled in a tiny corner of it.What an order this is!He was just dumbfounded. The first computer simulations were just the beginning.Kaufman still doesn't know why loosely connected networks behave so miraculously, leading him to think entirely new about genes and embryonic development.Using the original program as a template, he recreated countless computer simulations with some necessary modifications.When, he wondered, will order emerge?Why did it appear?And how did he test the theory with actual data? The most striking prediction of the model, he thought, was that true genetic networks must be loosely connected; tightly connected networks don't seem to be able to settle down into stable cycles.He didn't expect the real situation to be like his model, where there are only two types of input information for each gene, which is never the case in nature.But from his computer simulations and numerous calculations, he knew that the connection must be statistically loose.And when you look at that pile of data, man, the real web seems to be that loosely connected! So far so good.Another way to test the theory is to take a set of regulatory genes in a particular organism and see how many cell morphologies they produce.Because of Kaufman's deliberate attempt to study the typical behavior of networks, he knew he couldn't say definitively, but he could find statistical correlations.His hypothesis was that a cell morphology corresponds to a cycle in a steady state.So he started running larger and larger computer simulations, tracking how many loops would occur as the sample network scaled up.When simulating a network of more than 400 genes, he has determined that the number of cycles is roughly equal to the square root of the number of genes in the network.At the same time, whenever he was free, he spent time in the library of the medical school, searching all kinds of obscure reference materials, looking for comparable data in real organisms.When he finally pieced it all together, the answer came: indeed, the number of cell forms in an organism is roughly equal to the square root of its number of genes. God!It really worked!Kaufman said it was the most beautiful experience of his life.還沒讀完醫科二年級，他花在電腦上的錢已經好像流水一樣，但是他眉頭都沒有皺一下。 尋求外援 一九六六年，考夫曼念醫科的第二年，他寫了一封信給麻省理工學院的神經生理學家麥克古洛荷（Warren McCulloch），說明他已經完成了個遺傳網路模型，問麥克古洛荷有沒有興趣看看。 考夫曼承認，寫那封信有點魯莽。也是醫科出身的麥克古洛荷是神經生理學的大老之一，在電腦科學、人工智慧及心靈哲學方面也聲名卓著。一九四三年，他和一個叫比茲（Walter Pitts）的十八歲數學家共同發表了一篇論文。論文中，麥克古洛荷和比茲聲稱，可以用及、非、或等邏輯運算網路來模擬頭腦的操作。在當時，這是革命性的構想，而且發揮了極大的影響力。麥克古洛荷和比茲的模型不只是現在所稱的神經網路的第一個範例，也是科學家首度嘗試把精神活動看成資訊處理的形式來理解這種見解激發了後來人工智慧和認知心理學的發展。他們的模型也首次顯示出，非常簡單的邏輯閘門網路可能表現出異常複雜的運算，這個觀點也很快被納入計算機的一般理論中。而過去二十年，麥克古洛荷和一群忠實的信徒一直想辦法在他一九四三年提出的原始構想中，鑽研出其他意義。 不管麥克古洛荷是不是大老，他似乎是唯一能分享考夫曼研究成果的科學家。麥克古洛荷是我所知道唯一作過很多神經網路研究的科學家，而顯然神經網路和遺傳網路基本上是一樣的。 除此之外，在這個時刻，考夫曼迫切需要外在的支持。醫學院對他而言，有好也有壞。他當然獲得很多過去在牛津念哲學時所渴求的真實，但是他讀得並不是很好。我想我暗地裏惱怒老是要聽別人告訴我做這個、做那個，他說：在醫學院裏，你要做的就是精通各種事實、精通各種診斷方法，吸收別人的寶貴經驗和智慧，然後執行正確的流程。儘管在行醫的過程中，我也得到一些樂趣，但是卻得不到我要的那種美，這不像是在探索魔鬼的祕密。 同時，考夫曼致力於探索遺傳網路之美，也頗令教授不以為然。在醫學院念書幾乎像在做苦工一樣，二十四小時的輪班，永無休止的要求。目的是要讓你認清楚：病人第一。你要清晨四點鐘起床，去做你必須做的事，這我倒不介意。但是有些醫學院教授自以為是醫學的守護神，認為如果你的態度不對，那麼你永遠不會是個真正的醫生。 考夫曼記得大三時教他外科的教授尤其如此。他認為我心不在焉，他說得也沒錯。他說：我還記得他說不管我期末考是不是拿A，學期總成績他都要給我D。我想我期末考考了個B，但是他還是給我D。 所以你可以想見我在醫學院的樣子，古怪而悶悶不樂，外科拿了個D，對我來說，這是一大打擊。我曾經是拿馬歇爾獎學金的優等生，在醫學院卻掙扎過關，還有個外科教授告訴我，說我是多麼失敗！ 迷霧中的臉孔 事實上，當時他生命中唯一的光采就是，他結婚了，娶了個叫白安琪（Elizabeth Ann Bianchi）的藝術系研究生。她是個義大利裔美國女孩，考夫曼在牛津念書時巧遇到歐洲旅行的她。他還記得：我當時為她拉開門，心裏想：哇！這女孩真漂亮。從此，我就都幫她拉門了。 但是，即使白安琪都對這網路的玩意兒抱持懷疑的態度。白安琪比我專心多了，考夫曼說：她對醫學很有興趣，她和我一起上解剖學及其他的課。但是她對遺傳網路的反應是：聽起來不錯，但這是真的嗎？對她而言，這網路虛幻不實。 就在這個時候，考夫曼收到麥克古洛荷的回音：整個劍橋都為你的發現而雀躍。考夫曼回想起這件事，還覺得好笑，差不多一年後，我才知道當麥克古洛荷這麼說的時候，意思不過是他已經讀了我寄給他的報告，覺得還蠻有趣的。 不過當時，麥克古洛荷的回答令他又興奮、又驚訝。這遠超出他的預期。他壯起膽來回了一封信，解釋加州大學舊金山分校鼓勵醫科三年級學生到其他學校三個月，以吸取不同的經驗。他能不能申請到麻省理工學院，和麥克古洛荷一起作研究？ 麥克古洛荷回答，當然可以。而且，考夫曼和白安琪可以住在他家。 他們立刻接受邀請。考夫曼永遠不會忘記他第一次見到麥克占洛荷的情形。那是在一個冬夜九點鐘左右，考夫曼和白安琪在劍橋陌生而黑暗的街道上繞了又繞。大老遠開車跑來後，居然絕望的迷失了方向。然後，在迷霧中卻隱約出現了麥克古洛荷留著鬍鬢的臉孔，迎接我們到他家去。當麥克古洛荷太太端出乳酪和熱茶招待兩個筋疲力盡的旅客時，麥克古洛荷打電話給麻省理工學院的人工智慧大師明斯基（Marvin Minsky）：考夫曼來了。 良師益友 麥克古洛荷是個虔誠的教友派信徒，同時也是個體貼而迷人的主人。他擁有謎樣、如詩的風格和一顆自由漫遊於浩瀚知識大海的心靈，他對於探索思想的內在活動，總有永無止境的熱情。他的寫作頗有古風，科學論文旁徵博引，充斥著從莎士比亞到聖布納芬杜拉（Saint Bonaventura，十三世紀義大利哲學家、作家及樞機主教）的智慧話語，還取些像幻想從何而生？為什麼心靈存在於頭部之中？穿過玄學家的私室之類的題目。他喜歡猜謎語和雙關語。而且，他是世界上少數能夠講贏考夫曼的人。 麥克古洛荷常常會把你陷進冗長的討論中。考夫曼說。麥克古洛荷習慣在考夫曼洗澡時，跟著他進浴室，把馬桶蓋翻下來，在考夫曼忙著輕洗耳朵上的肥皂時，坐在那裏快樂的大談網路和不同種類的邏輯功能。 然而，最重要的是，麥克古洛荷成為考夫曼的導師和朋友，正好像他過去對待其他學生一樣。麥克古洛荷知道考夫曼到麻省理工學院的目的，是要作龐大的電腦模擬，以便為他所研究的網路行為蒐集詳細的統計資料。麥克古洛荷介紹他認識明斯基及明斯基的同事派普特（Seymour Papert），他們安排考夫曼用當時稱為MAC計畫（MAC代表machine︱aided cognition，機器輔助認知）的強力電腦來作電腦模擬。麥克古洛荷同時還安排一個精通電腦語言的大學生為考夫曼寫程式，結果，他們為上千個基因作電腦模擬。 同時，麥克古洛荷還介紹考夫曼認識理論生物學小而熱情的世界。就是在麥克古洛荷的家中，考夫曼見到了神經生理學家軻文，軻文在一九五○年代末期和六○年代初期曾經擔任麥克古洛荷的研究助理。考夫曼認識他的時候，他剛受命重振芝加哥大學的理論生物學小組。考夫曼也是在麥克古洛荷的辦公室認識了英國薩西克斯大學（University of Sussex）的古德溫（Brian Goodwin），從此就結成莫逆之交。 麥克古洛荷就像高中英文老師托德一樣，他是第一個認真把我當年輕科學家看待，而不是只把我當成學生的人。考夫曼說。令人悲傷的是，麥克古洛荷在幾年後（一九六九年），就逝世了。但是，考夫曼仍然有一點覺得自己繼承了麥克古洛荷的志業。麥克古洛荷把我引進一個我再也不曾離開過的學術世界。 indeed so.去麻省理工學院之前，考夫曼已經決定畢業後要投身科學研究，而不去行醫。但是，麥克古洛荷介紹他認識的這群科學家，才真正把他引進這個圈子。 經由軻文、古德溫及其他人的介紹，我才在一九六七年受邀參加生平第一個科學研討會。He said.這個研討會是由英國胚胎學家威丁頓（Conrad Waddington）所召開的一系列理論生物學研討會的第三場。在一九六○年代，這些會議嘗試要做的事正像今天的聖塔菲研究院。考夫曼說。真是太棒了。從清晨四點鐘起來抽血，檢查大便樣本（親手接觸現實！），一變而為飛到北義大利的湖邊別墅，周圍盡是令人訝異的人。史密斯（John Maynard Smith）在那兒，湯姆（Rene Thom）剛發明了災難理論（catastrophe theory），芝加哥的路翁亭（Dick Lewontin）、列溫斯（Dick Levins），倫敦來的渥普特（Lewis Wolpert）都在那兒，這些人直到現在都還是我的朋友。 所以我發表演講，談到遺傳網路、細胞種類等等。之後，我們到陽台上喝咖啡。軻文走出來問我願不願意去芝加哥作研究，我幾乎不加思索就回答：當然願意！足足有一年半，我都沒有問軻文，我的薪水會是多少。 Different routes lead to the same goal 亞瑟到聖塔菲研究院的第一天中午，就和考夫曼沿著峽谷路上的美術館，漫步到考夫曼最喜歡的水洞。此後的兩個星期，幾乎每天他們都一起碰面吃中飯，或只是散散步。 他們多半邊走邊談，考夫曼幾乎比亞瑟還喜歡戶外空氣。十幾歲、還在當童子軍時，考夫曼就參加過不計其數的遠足和露營活動，上大學以後，又迷上了滑雪和登山，現在只要有空，他仍然常常徒步旅行。所以，當考夫曼和亞瑟漫步於峽谷路上，或是從修道院後面走上山，在山頂遠眺聖塔菲全景及連綿的山脈時，他們談了很多。 亞瑟開始發現，考夫曼有一種難以言喻的傷感。偶爾，他會在講笑話、猜字謎、展現他無所不在的好奇心、或滔滔不絕的談他的點子時突然停頓，臉上閃過一絲哀傷。一天晚上，當亞瑟夫婦和考夫曼夫婦一起出外晚餐時，考夫曼吐露了他的心事：一九八五年十月的一個星期六晚上，考夫曼和白安琪回家的時候，發現他們十三歲的女兒被車撞了，情況很嚴重，正躺在當地的醫院裏。他們和兒子一起衝到醫院，卻發現女兒已在十五分鐘前過世了。 在事情已經發生了五、六年後的今天，考夫曼已能夠鎮靜的講完整個故事。但是，那天晚上他卻情不自禁，他一向寵愛這個女兒。我的心碎了，痛苦的不得了。我們上樓去看她，女兒殘破的軀體躺在那裏，逐漸冷卻。我受不了。那天晚上，我們三個人躺在一張床上，抱頭痛哭。她脾氣不太好，但是有一種不尋常的聰慧，我們都覺得她是我們四個人中最出色的一個。 他們都說，時間會治療一切，但事實上並不盡然。只不過悲傷不再那麼常爆發而已。考夫曼說。 當他們在聖塔菲山間漫遊時，亞瑟禁不住被考夫曼的秩序和自我組織的概念所吸引。諷刺的是，當考夫曼提到秩序時，他的意思顯然和亞瑟所說的雜亂一樣，也就是指複雜系統永無休止的自我組織成形態的衝動。不過，考夫曼和亞瑟用的形容詞完全相反，這並不足為奇，因為他們的觀念起源完全南轅北轍。亞瑟談到雜亂，是因為他的出發點是冰冷而抽象的經濟均衡世界，在這個世界裏，市場法則像物理法則一樣精確無誤的決定一切。考夫曼談到秩序，是因為他的出發點是達爾文雜亂而不定的世界，在那個世界裏，沒有法則可言，意外和天擇決定一切。但是，儘管出發點完全不同，他們卻獲致相同的結論。 同時，考夫曼也為亞瑟的報酬遞增所迷惑。我一直沒辦法明白為什麼這是個新觀念，生物學家研究正回饋已經很多年了。He said.他花了很多時間才明白新古典學派的世界觀是如何的靜止不變。 你的問題正是我的問題！ 還有個一直困擾亞瑟的經濟問題就是技術變遷，考夫曼聽到後更迷惑了。技術變遷已經變成政治上最炙手可熱的議題。你可以在隨手拿起的每一份報紙和雜誌上，感覺到一股潛在的焦慮：美國還有競爭力嗎？我們是不是已經喪失了傳說中的美國創造力？日本人是不是會一個產業接著一個產業的，把美國人打垮？ 這些問題都很好。亞瑟解釋給考夫曼聽，問題是，經濟學家沒有答案，至少基本經濟理論提不出解答。技術發展的動態就像黑盒子一樣。直到十五或二十年前，大家都還認為技術是憑空從天上掉下來的，是像愛迪生之類聰明的發明家躺在浴缸中靈光一閃而產生的。嚴格來說，技術根本不算經濟學的一部分，技術是外來的，是由非經濟過程神奇地孕育出來的。近年來，有些人嘗試把技術看成是由經濟體系內在產生的，但是他們通常把技術看成投資於研究發展後的成果，幾乎就像商品一樣。亞瑟認為儘管其中有一部分事實，卻沒有搔到癢處。 他告訴考夫曼，打開經濟史，而不要單看經濟理論，你會發現技術並不全然像商品一樣，反而更像演化中的生態系統。尤其是，創新絕少在一片真空中誕生，既存的創新技術往往帶動了新的發明。例如，雷射印表機基本上就是個雷射影印機，再加上電腦線路來告訴影印機該印什麼。所以當電腦技術、雷射技術和影印技術都發展成熟時，才會有雷射印表機。但是，也惟有在人們需要精巧、高速的列印技術時，才有可能發生。 簡單的說，技術形成了緊密相連的大網，或是套用考夫曼的字眼網路。更重要的是，技術網非常的活躍而不穩定，幾乎能像有機體一樣成長。就好像雷射印表機帶動了桌上排版軟體的發展，桌上排版又為繪圖程式打開一片新的天空。甲、乙、丙技術可能帶動丁技術，以此類推，於是所有可能的技術連結成一個網路，隨著相關技術的發展而日益成長，因此經濟也就變得更複雜。亞瑟說。 技術網還會更進一步爆發新的創造，或是經歷大滅絕，就好像生態系一樣。舉例來說，像汽車這樣的新技術出現，取代了舊技術馬匹。和馬匹一起被淘汰的有鐵匠店、快速馬車、水槽、馬廄、照顧馬匹的工人等等，依賴馬匹的技術次網路在經濟學家熊彼德所稱毀滅的狂風下，驟然崩潰。新商品和服務的嶄新網路開始發展，紛紛填補了過去的商品和服務所開發的利基市場（niche）。 亞瑟說，這種過程正是報酬遞增的絕佳範例。一旦新技術開始為其他的商品和服務開創新的利基，掌握了利基市場的人也會有強烈的誘因來協助新技術的發展。更重要的是，這種過程是鎖定現象的主要驅動力，由某種技術所帶動的利基市場愈多，就愈難改變這種技術，除非出現了遠勝過現有技術的新技術。 所以，技術網的想法和他的新經濟觀息息相關。問題是，他所發展出來的數學方程式一次只能適用於一種技術，而他真正需要的是像考夫曼所發展出來的網路式模型。他問考夫曼：你能不能建立一種模型，其中技術一旦創新，也就同時被開啟了？ 考夫曼驚愕的聽著亞瑟的長篇大論。儘管所用的語言不同，亞瑟所形容的問題正是考夫曼花了十幾年時間研究的問題。幾分鐘之內，考夫曼開始向亞瑟解釋，為什麼技術變遷的過程完全就好像生命的起源。 找到知識的天堂 一九六九年，差不多就在考夫曼抵達芝加哥的理論生物學小組的時候，他第一次有了這個構想。 他說，讀了幾年醫科以後，他覺得芝加哥好像天堂一樣。事實上，回顧以往，芝加哥是第二個令他興奮的知識殿堂。這是個不尋常的地方，充滿了不尋常的能人，他說：就像我在義大利研討會認識的那群朋友。軻文正從事他在皮層組織的突破性研究以簡單的方程式來形容穿過腦部神經細胞的刺激和抑制波。史密斯也在作突破性的演化動力學研究，他以一種叫博弈理論（game theory）的數學技巧，來澄清物種間競爭與合作的本質。當時史密斯因為在薩西克斯大學休教授年假，而來到芝加哥，他協助考夫曼解決了很多為網路作數學分析的困難。 身邊環繞著同事和知音，考夫曼很快就發現他不是唯一想到網路統計特性的人。例如一九五二年，英國神經生理學家艾胥彼（Ross Ashby）在他的書腦部設計（Design for a Brain）中，就思考了同樣的問題。關於複雜網路的遺傳行為，他問了一些很類似的問題，我卻完全不曉得。考夫曼說：我一發現這件事，就立刻與他聯絡。 同時，考夫曼發現在發展遺傳網路的時候，自己已重新發明了當時物理和應用數學上一些最前衛的研究。他的遺傳調節網路結果變成物理家稱之為非線性動力學的特殊案例。從非線性的角度來看，事實上，很容易就可以了解，為什麼連結鬆散的網路能輕易的自我組織成穩定的循環。在數學上，這道理就和落在山坡上的雨水流入谷底湖中一樣。在所有可能的網路形態中，穩定的循環就好像盆地。 和這些網路艱苦奮鬥了六年之後，考夫曼感到很滿足，他終於開始清楚箇中道理，但是，他還是禁不住覺得好像還缺了些什麼。討論遺傳調節網路的自我組織很有趣，但是在分子的層次，遺傳活動要依賴極其複雜的RNA（核糖核酸）、DNA分子。RNA和DNA又是如何誕生的呢？ 生命是怎麼開始的？ 根據生物學教科書的標準理論，生命的起源很直接。DNA、RNA、蛋白質、多醣類，以及其他所有生命的分子，一定是在幾十億年前誕生在某個溫暖的小池塘中，小池塘中同時也從原始大氣中聚積了胺基酸之類的生命基本單位。事實上早在一九五三年，諾貝爾化學獎得主游理（Harold Urey, 1893︱1981）和他的研究生密勒（Stanley Miller, 1930︱）就以實驗證實，只需要偶爾出現的閃電提供化學作用的能量，最初大氣中的甲烷、氨和其他類似的氣體就能自發的產生生命的基本單位。後來的說法就變成：在湖中或小池塘中形成的這些簡單化合物，經過進一步的化學作用之後，變得愈來愈複雜。最後，產生了包含DNA雙螺旋結構及它的單股堂兄弟RNA的分子集合體。DNA和RNA都有繁殖的能力。一旦生物開始有了繁殖能力，接下來的一切就遵循物競天擇的法則。這就是標準理論的說法。 但是，考夫曼不接受這個說法。不談別的，生物分子是個龐大的架構，例如，要製造一個蛋白質分子，你可能就必須依照精準的次序連結起幾百個胺基酸。即使在現代實驗室中，應用所有最先進的生物科技工具，都很難製造出像這樣的結構，那麼，這麼複雜的東西怎麼有可能在池塘中自我形成呢？很多人嘗試計算可能發生的機率，答案都差不多：如果生命真是隨機形成的，那麼單單製造一個有用的蛋白質分子，所花的時間就要比宇宙的歷史還久，更不要談製造一個可以完全運作的細胞所需要的，數不清的蛋白質、糖類、脂肪、核酸了。即使你假設在我們可觀測到的宇宙中，數百萬的銀河系裏數以兆計的星球中，都有像地球這樣擁有溫暖海洋及大氣層的行星存在，它們能產生生命的可能性仍然微乎其微。如果生命的起源真的是種隨機事件，那麼，這的確是個奇蹟。 說得更精確一點，考夫曼不相信標準理論，是因為標準理論把生命的起源和DNA的出現畫上等號。對考夫曼而言，生命的起源要依賴這麼複雜的物質出現，顯然不合理。沒錯，DNA能自我複製，但是關鍵在於它必須先解開兩股螺旋，然後進行自我拷貝，過程中還要依賴一群扮演協助者角色的蛋白質分子。這一切怎麼可能都發生在小池塘中呢？ 我有一股想一探究竟的衝動，就像當初我想發現遺傳調節網路的秩序一樣，考夫曼說：DNA太奇妙了，我只是不相信生命的起源完全依賴於如此特殊的物質，我的說法是如果上帝當時賦予氮不同的原子價（DNA分子中充滿了氮原子），是不是還會有生命？就我看來，把生命看成如此微妙的平衡，是個驚人的結論。 分子的月下老人 但是，考夫曼想，誰說生命最重要的物質是DNA？誰說生命的起源是隨機事件？也許能自我複製的系統還有另外一種誕生的方式，一種能從簡單的反應自行衍生成生命體系的方式。 好吧，想想有著小小的胺基酸、糖類等物質活躍其間的太初渾湯（primordial soup）是什麼樣子。顯然，你不可能期望這些物質就自然形成一個細胞，但是你能想像它們彼此有些隨機的反應。儘管這些隨機的反應不會產生太神奇的東西，但是只要計算就知道，一般來說，它們都會產生一些有短鏈和分支的小分子。 單單如此並不是就比較可能產生生命。但是，考夫曼假設有些在太初渾湯中到處漂浮的微小分子，能夠扮演觸媒的角色充當顯微鏡下都看不見的月下老人。化學家對這種現象再熟悉不過了：有催化作用的分子逮住兩個匆匆經過的分子，為它們牽線，讓它們很快起化學作用而結合，然後觸媒分子就放開這對新婚夫婦，逮住下一對，如此這般的繼續下去。化學家也知道還有一些觸媒分子扮演刀斧手的角色，側身挨近一個又一個的分子，把它們斬成兩半。兩種觸媒都是現代化學工業的中流砥柱，如果沒有觸媒，幾乎不可能產生汽油、塑膠、染料和藥劑。 考夫曼想，好，想像現在你的太初渾湯中有一些分子A正忙著催化分子B的形成。分子A可能不是非常有效的觸媒，因為基本上它是隨機形成的；但是它不需要非常有效，因為即使是微弱的催化作用都能讓B分子形成的速度比原先快。 現在，假設分子B也有微弱的催化作用，於是分子B又為分子C催生。假設分子C也是觸媒，以此類推。他推論，如果這鍋太初渾湯夠多，而且一開始分子的種類夠多，那麼很可能在某個時候就出現了能催生分子A的分子Z，形成一個循環。但是現在分子A愈來愈多了，也就是說能催化分子B的觸媒愈來愈多，因此也就會有更多的分子C被催生，以此類推。 換句話說，如果太初渾湯中的條件正確的話，你根本不需要等待隨機的化學作用。太初渾湯中的化合物會形成有一貫性而且能自我強化的化學反應網。而且，在這個反應網中的每一個分子都能促進其他分子的形成，所以，反應網中的所有分子都會穩定的比網外的分子更快速成長。簡言之，一個反應網是一個自動催化組（autocatalytic set）。 當考夫曼了解這一切時，他肅然起敬。秩序再一次出現了，自然形成的秩序！秩序從物理和化學法則中自然產生，秩序從分子的混沌中自動顯現，而且形成一個不斷成長的體系。真是太美了，筆墨都難以形容。 但是，這是生命嗎？ 不是，考夫曼必須承認，如果你是指今天所謂的生命，那麼它不是生命。自動催化組中沒有DNA、沒有遺傳密碼、沒有細胞薄膜，事實上，它只不過是漂浮在古老池塘中隱隱約約的一群分子，並不是真的獨立存在。如果外太空有個達爾文恰好經過，他都很難察覺有什麼不尋常，自動催化組中的每一個分子看起來和其他分子幾乎一模一樣，組織的本質並不存在於個別的分子中，而是存在於整體的活動（集體行為）中。 大自然的衝動 深一層看，也許自動催化組有生命，它會顯現一些非常近似生命的特性，例如它能成長。基本上，自動催化組沒有理由不能逐漸產生愈來愈多的分子，並且所製造的分子愈來愈複雜。而且，它可能還有某種新陳代謝的功能：反應網的分子會穩定的吸收食物分子，那是在太初渾湯中隨處漂浮的胺基酸和其他簡單的化合物。經過催化作用相互黏結後，就會形成更複雜的化合物。 自動催化組甚至會顯示出一種原始的繁殖方式，如果小池塘中一個自動催化組恰巧因為洪水之類的原因，而流到鄰近的小池塘中，它會立刻在新環境中開始生長。當然，如果另外一個不同的自動催化組已經在池塘中，兩個分子組織就會爭奪資源，物競天擇的大門立刻敞開，展開去蕪存菁的過程。可以想見，經過天擇過程而存活的分子組織不是最強壯、最能適應環境變遷的，就是擁有最多有效的催化分子，或是具有最複雜結構的分子組織。事實上，可以想見，最後去蕪存菁的過程促成了DNA及其他物質的誕生。真正的關鍵在於先產生能存活而且能繁殖的物質，之後，演化很快就會發揮功效。 好，這個結論是建築在許多的假設上。但是對考夫曼而言，這是到目前為止關於生命起源最可信的解釋。如果這是真的話，就表示生命的起源不必等待幾乎不可能的機會來產生一組極端複雜的分子，生命的確是由簡單的分子自動形成的。也就是說生命不必然是偶發的意外，而是大自然自有一種不可抗拒的衝動，要永無休止的進行自我組織。 考夫曼著迷了，他立刻投入繁複的計算、電腦模擬和隨機網路模型中，把在柏克萊做過的事情再重複一遍。他希望了解自動催化組的自然法則。他想，好，你不清楚遠古時代到底有些什麼化合物，以及它們起了什麼化學反應，但是，至少可以想想可能性。形成自動催化組的可能性很低嗎？還是幾乎不可避免？看看數據吧！假定你有幾種不同的食物分子，如胺基酸等等，又假定在太初渾湯中，食物分子開始連結成聚合物鏈。以這種方式，你可以形成幾種不同的聚合物？在這些聚合物中，可能在發生多少反應後，才能產生一個巨大的反應網？這個巨大的反應網自成一個循環、形成自動催化組的可能性又有多大？ 考夫曼說：當我整個想一遍後，我發現顯然化學反應的數目要比聚合物數目成長得更快。所以只有當每一聚合物所能催化的反應數目達到某個固定值時，才會有複雜顯現，也才能進一步出現相互的自動催化。換句話說，這就像他的遺傳網路一樣：如果太初渾湯通過了某個複雜性的門檻，那麼就會經歷這種滑稽的相變，幾乎不可避免的就產生了自動催化的組織，而生命形貌也自然誕生。考夫曼覺得這整件事都太美了，一定就是這樣。一直到現在，我都對這個景象深信不疑，我相信這就是生命的起源。 跨越臨界點 亞瑟也深信不疑。他覺得這個想法很棒，不只是因為這是關於生命起源的絕妙想法，而是自動催化作用和經濟學相像得不容忽視。他和考夫曼那幾天無論在山間散步，或在水洞午餐，都反覆討論這個想法。 他們都同意，自動催化組是分子之間的轉換網，正如同經濟是商品和服務的轉換網一樣。事實上，自動催化組幾乎就等於一個極其微小的經濟體系吸取原料（原始的食物分子），再把原料轉換成有用的產品（更多的分子）。 而且，自動催化組能自行演化，也就是時間愈久，就變得愈複雜，這和經濟體系完全一樣。這是最讓考夫曼著迷的地方。如果融合舊科技能導致創新，那麼當可獲得的技術愈來愈多時，可能的創新也就會快速增加。事實上，一旦你跨越了某種複雜性的門檻，經濟就會產生像自動催化組一樣的相變。技術還沒有到達相當複雜度的國家，因為只依賴少數重要產業，經濟體質是脆弱而停滯的。在這種情況下，這個國家有多少投資並不重要。如果你所做的只是生產香蕉，那麼除了生產更多香蕉以外，不會發生別的事情。但是，如果一個國家想辦法多角化、提高技術複