Chapter 126 20.5 Self-regulating living systems

Stuart Kaufman's simulation was like any mathematical model: rigorous, novel, and garnering the attention of scientists.Maybe more than that, because he was simulating a hypothetical network with a real (computer) network instead of, as usual, a real network with a hypothetical network.Nevertheless, I admit that this is only a small step in the long journey of applying the abstract concepts of pure mathematics to irregular realities.Nothing is more irregular than online networks, biogenetic networks, and international economic networks.However, Stuart Kaufman was very eager to extrapolate the results of his general experiments to real life.The comparison between complex real-world networks and his own mathematical simulations running on silicon cores is the holy grail Kaufman has been scrambling for.He considers his models "as if they were real".He bets that swarm networks behave similarly at some level.Kaufman likes to say, "IBM and E. coli don't see the world differently."

I am inclined to believe his point of view.We have technology that connects everyone to everyone else, but some of us who try to live that way find that no matter what we are trying to accomplish, we are disconnected.We live in an era of accelerated connectivity, in fact, steadily climbing Kaufman's hill.But it's hard to stop ourselves from climbing over the top and sliding down the slopes of increasingly connected and less adaptable.Disconnection is the brake, it can prevent the system from being over-connected, and it can keep our cultural system on the edge of the highest degree of evolution.

The art of evolution is the art of managing dynamic complexity.It's not hard to connect things, but the art of evolution is to find organized, indirect, limited ways to connect. Kaufman's Santa Fe Institute colleague Chris Langton derived an abstract property, called the lambda parameter, from his experiments with swarm models of artificial life. The lambda parameter can predict the possibility of a group producing a "best balance point" of behavior under a certain set of rules.Systems outside this equilibrium point tend to get stuck in one of two modes: they either freeze at a few lattice points, or they scatter into white noise.Those values ​​that fall within the sweet spot range give the system the longest period of meaningful behavior.

By tuning the lambda parameter, Langton can tune the world to make it easier to learn or evolve.Langton called the critical point between the changing state between several fixed points and the phaseless gas state a "phase transition"—the same term physicists use to describe the transition from liquid to gas, or from liquid to gas. solid.Most surprisingly, however, Langton found that the lambda parameter slowed down as it approached the phase transition—the "sweet spot" of maximum fitness.That is, the system tends to stop at this edge and not overrun.As it approaches this extreme point of evolution, it becomes wary.Langton likes to picture this as a system surfing on a slow-moving, never-ending perfect wave, with time running slower the closer it gets to the top of the wave.

This deceleration at the "edge" is key to explaining why unstable embryonic living systems can continue to evolve.When a stochastic system approaches a phase transition, it is "pulled" toward and docked at the optimal equilibrium point, where it evolves and seeks to retain that position.This is the self-static feedback loop it builds for itself.Since the sweet spot is hard to describe statically, perhaps it would be better to call this feedback loop "autodynamic". Stuart Kaufman also talked about "adjusting" the parameters of his simulated gene network to the "best balance point".There are also countless ways to connect millions of genes or neurons, but outside of the ways of connecting, some settings with a smaller number are much more important to promote the learning and adaptation of the entire network.A system at this evolutionary balance can learn the fastest and evolve most easily.If Langton and Kaufman are right, an evolving system will find this balance on its own.

So how did all this happen?Langton found some clues.He found that this point was on the edge of chaos.He argues that the most adaptive systems are so uninhibited that they are only a hair's throw away from being out of control.Life is neither a stagnant system with no communication nor a deadlocked system with too much communication.Life is a living system tuned to the "edge of chaos" - right at that lambda point, where the flow of information is just enough to keep everything teetering on the edge. A little looser rein can make a rigid system work better; a little more organization can improve a disorganized system.Mitch Waldrop explains Langton's concept this way in his book "Complexity": If an adaptive system is not running satisfactorily on the right track, unselfish efficiency will push it towards the optimal equilibrium point .If a system rests on the pinnacle of balance between rigidity and chaos, then if it strays from its original position, adaptation will pull it back to the edge. "In other words," Waldrop writes, "learning and evolution will stabilize the edge of chaos"—a self-reinforcing sweet spot.We should perhaps say that it is dynamically stable, since its position is constantly changing.Lynn Majliss calls this changing, dynamically constant state "fluid stability"—that is, clinging tightly to a moving point.It is the same persistent teetering that keeps the chemical pathways of Earth's biosphere purposefully out of balance.

Kaufman called the system established within the range of λ a "suspension system".These systems hover at the juncture of chaos and rigid order.Suspended systems can be seen everywhere in the universe, even outside the biosphere.Many cosmologists, such as John Barrow, believe that the universe itself is a suspended system, reaching an unstable equilibrium over a series of very precise values ​​(such as the parameters of gravity or the mass of electrons).If these values ​​were changed by a tiny bit, even by an insignificant one-in-a-billionth, the universe could have collapsed in the first place, or could not have formed at all.There are so many such "coincidences" that I could write several books.According to the mathematical physicist Paul Davies, these coincidences "take together to provide powerful evidence that life as we know it is so sensitively dependent on the pattern of physical laws, on the fact that what appears to be chance is actually Nature chooses values ​​for various particles and interactions." Simply put, the universe and life we ​​touch hovers on the edge of chaos.

What if the suspension system could adjust itself independently of the creator?For a self-balancing complex system, it will gain a huge evolutionary advantage in the biological sense.It evolves faster, learns faster, and adapts more easily.If evolution has selected for self-regulatory functions, then "the ability to evolve and adapt may itself be a major evolutionary achievement," Kaufman said.And self-regulation is an inevitable choice for higher evolution.Kaufman proposes that the genetic system indeed regulates itself for optimal flexibility by adjusting factors such as the number of connections and the size of chromosomes within its system.

Self-regulation may be the magic key to never-ending evolution, the holy grail of open evolution.Chris Langton formalized open evolution as a system that constantly adjusts itself to increasing complexity, or in his conception, as a Keep a balanced system. In Langton and Kaufmann's framework, nature begins as a collection of interacting aggregates that autocatalyze to form new aggregates and connect into networks that allow evolution to take place to its maximum extent.This evolution-rich environment gave rise to cells, which learned to tune their internal connections to keep the system in an optimal evolutionary state.Every step taken on the edge of chaos is carefully treading the path of optimal flexibility, constantly increasing its complexity.As long as the system is on this evolutionary crest, it will keep rushing forward.

What we want in man-made systems is something similar, Langton said.The first goal any system seeks is survival.The second is the ideal parameters to ensure the maximum flexibility of the system.But the most exciting is the third level of goals: finding strategies and feedback mechanisms for systems to continuously enhance self-regulation during evolution.Kaufman hypothesizes that if systems are built to self-regulate, then they adapt easily, meaning they are the inevitable target of natural selection.Being able to tap into natural selection would be one of the preferred abilities. When Langton and his colleagues were searching for the best balance of life in the space of possible worlds, I heard them say that they were surfing the endless summer, looking for the perfect and slow wave.

"What I'm looking for is something that's a window paper away from me," Ricky Bagley, a researcher at the Santa Fe Institute, told me.He further explained that it is neither regular nor chaotic; on the verge of being out of control and dangerous. "That's right," replied Langdon, who overheard our conversation. "It's really like the lapping waves, they're pounding against the shore, like a heartbeat. And then all of a sudden, wow — there's a big wave. And that's what we're all looking for."