Skynet Is Us
How biology is more sophisticated than Hollywood scriptwriters
On Sunday 7th June as I was walking up through Lausanne on the way to the gym, I saw evolution in action.
I didn’t realize the fact immediately, because from ground-level it’s not obvious when something radically new is occurring even when it’s happening right in front of our eyes.
Here was the obvious phenomenon: a large group of earnest people all dressed in black were congregating in the Place de Riponne. They were eagerly virtue-signaling, waving placards demanding an end to police brutality.
I was initially puzzled by this, because although no nation’s police force is one-hundred-percent perfect, the Swiss police come pretty close, as do those of the Scandinavian countries, the Netherlands, and Germany. The last seemingly overt case of racist-inspired police aggression in Switzerland occurred back in 2014. So protesting against police racism and brutality in Switzerland is a bit like people in Iceland protesting about the scorching hot winters they have to endure, or Trump supporters complaining that their average IQ is too high.
Perhaps these earnest folk were protesting against police brutality in the USA, but not a single sign indicated that. Nor would such a protest have made any sense because no one in the USA is going to pay any attention to a demonstration in Switzerland. Indeed, 95% of US citizens can’t even locate Switzerland on a map. So on an a priori basis the demonstration made no sense.
And then I realized the phenomenon had nothing to do with the nominal cause at all. I realized I was seeing evolution occurring in front of me.
To explain what I mean, we need to step back to a quick sketch of the origins and subsequent development of life on Earth.
While the Victorian definition of life went something along the lines of “it respires and it reproduces” it’s become apparent that this is a very poor basis for understanding anything. A virus doesn’t respire but no one could claim that a virus can’t reproduce. So perhaps a more adequate definition of life is a self-organizing physical system that propagates.
If we were to fill a swimming pool with basic molecules and leave it for a sufficient period of time, we’d return to discover that complex molecules have spontaneously formed by statistical probability alone. We’d see amino acids and benzine rings and lipids: all the building-blocks required for living cells. It seems that life emerges spontaneously from the molecules that existed on Earth over four billion years ago. Give the right molecules enough time and eventually a tiny fraction of a percent form sustainable configurations. And that tiny fraction of a percent then reproduce (by splitting in two) and end up being very significant indeed.
Moving up the evolutionary ladder a couple of rungs, we see single-celled organisms that sometimes cooperate in vast numbers. Bacterial films are one example of this kind of multi-cellular cooperation on the most basic level. No one’s quite sure how eukaryotic life got started (my suspicion is a defect in the bacterial splitting process that left two cells joined, which then split again to form four joined cells, and so forth), but being multi-cellular opened up a dramatic range of new possibilities. Even something as tiny as a nematode worm a mere half-millimeter in length has around a thousand cells all cooperating to enable the collective to survive and reproduce. None of the cells individually has any idea what’s going on; the cooperation is purely a mechanical process.
We see something similar with social insects. A handful of ants on their own wander around chaotically and have limited survivability but a million ants together reliably form self-organizing colonies that can sustain themselves for decades. There’s no Central Planner Ant in charge of everything, nor are there self-important Ant Committees deliberating over what to do for the greater good. Ants are automata, operating on the basis of a few simple hardwired rules which are, en mass, enough to ensure coherent group behaviors.
If we think about our own bodies, the same rule applies. Our kidney cells aren’t sitting there all chanting “we’re so glad we’re kidney cells, working hard for the greater good!” None of the cells in our body, even the neurons in our brain, has any self-awareness. We only achieve our limited illusion of consciousness in consequence of vast numbers of neurons, each with up to ten thousand connections, all interacting in a manner that permits us to function in relatively complex ways. And perhaps one or two people in a thousand, by natural genetic variability, will even be capable of moments of coherent rational thought.
The really interesting thing about the brain is that there’s no massive blueprint stored away in our DNA that tells how to wire up each neuron. Instead, there are a few simple rules that neurons follow and the result is a coherent network. In other words, our brains self-organize in much the same way as ant colonies. With sufficient neurons and sufficient ants, a coherent entity emerges.
Let’s stay with the brain a while, because the comparison is going to be illuminating.
Neurons interact with other neurons. The more interactions, the greater the complexity of response. Neurons have what’s called an action potential. When the action potential is reached, the neuron fires and thus stimulates other neurons in the network. Neurons interact to enhance or diminish each other’s action potentials, and thus complex signaling can emerge. As neurons connect, they “test fire” a few times to validate the connection; after that, connections are strengthened on the basis of the frequency with which they are used.
And now we step sideways to look at the Internet.
Today, most people on the planet have some sort of access to the Internet. It’s the first time in history since our distant ancestors evolved that nearly our entire species has been connected in time and space.
For the first decade or so the Internet was the purview of middle-class Western folk, often well-meaning liberals who thought they were witnessing the dawn of an electronic utopia. Then it became a wealth machine for a few global corporations. But over the last ten years, as it’s become increasingly ubiquitous, the Internet has started to resemble axons and dendrites with us humans as quasi-neuronal cell nuclei. And like all neurons, we began with a few test-firings.
Anyone familiar with social media will know all about the way in which transient fads spread through the Internet like wildfire. Today it’s planking, tomorrow it’s pouring ice over your head, the day after that it’s rushing to some location to take a selfie like the million other people you’ve just seen on your InstaSnap app. These were more-or-less random test-firings, meaningless in themselves but the beginnings of synaptic activity.
This year, with the global hysteria over covid-19, we’ve seen the first system-wide activity. Even if you didn’t rush out to get a tattoo or plank on your balcony or pour ice over your head, you got scared out of your wits by the incessant sensationalism that bombarded you whenever you looked at your hypnotic screen. And you couldn’t stop looking. You became a good little neuron, ready to respond to an incoming stimulus.
The anti-police-brutality protest in Lausanne had nothing to do with police brutality. It was just another firing of the neuronal network. We’ve started to lay down quite strong connections now, so it’s easier and easier to set off cascades of firing. These neural cascades aren’t purposeful (though of course to the people involved, there’s the reassuring illusion of purpose). The system hasn’t begun to self-organize yet as we’re still in the early stages. And numbers are small. Even if seven billion people are connected to the Internet and on average have 400 connections to others, this is a tiny number compared to the 86 billion neurons in the human brain, each capable of having up to 10,000 connections.
But it’s a start. The brains of rats have only around 21 million neurons with around a thousand connections each and rats can solve sophisticated problems. It’s impossible not to wonder how much sophistication is therefore available from a network of seven billion pseudo-neurons with an average of 400 connections each.
Just like the neurons in our own brains, we won’t be conscious of the larger system of which we each individually form a part. It’s likely that we’ll gawp at the resulting phenomenon entirely unaware of the fact that we’re some of the enabling components. Just as individual ants are incapable of comprehending the emergent properties of the ant colony in which they exist, it’s plausible to posit that we too will be systemically incapable of comprehending the emergent properties of the network in which we each play a tiny role. Our brains may simply not have the wiring necessary to grasp the larger picture, much as our individual neurons are incapable of understanding the system in which they exist.
Hence the reference to the pulp movie released in 1985, in which killer robots substituted for Hollywood’s former go-to bad guys (the ones who lived in tepees and wore feather headdresses).
In the world of sci-fi the future was all about a powerful computer that by means of networking with other computers around the world would become vaguely sentient and self-directing. Ironically, we’re the ones becoming part of a huge networked organism, the actions of which we as individuals may remain forever ignorant.
Of course, we’re still in the very early stages. Individual humans can to some degree control the messages that are sent across this vast network and thus influence the lives of billions of people who can’t help but believe whatever it is they’re told by purported sources of authority. We are, after all, a group species hardwired to do just that. But even here there are clear limitations: study after study has shown conclusively that we humans embrace simplistic ideas and shun complexity. Our ape-brains are hardwired to avoid thinking whenever possible because thinking burns glucose and for most of our evolutionary history that glucose would likely be needed for powering muscles. We had to run from predators and forage for food; thinking came far down the list of priorities.
So the signals that are carried by this quasi-neural network we call the Internet are today mostly exceedingly simple. But so are the signals in the brain of an ant, yet ant colonies can achieve sophisticated self-organization. The signals in our own brains are simple electrical impulses, yet can achieve sophisticated results. Just because we, the components, operate on the basis of extreme simplicity doesn’t mean that the emergent properties of such a network will be simplistic. Indeed, everything we know about biology tells us the converse is true.
We are seeing the emergence of a system that will eventually be quite different from simply being the sum of its component parts.
It’s not obvious what we should call such a system, but it does seem clear that it will have its own emergent properties and that it will have significant impact on the way things unfold from here on out. I suppose we’ll just have to hope that whatever emerges will be less clueless and less destructive than we ourselves have been.