Ways of Being — James Bridle on Seeing Past Human Intelligence

Ways of Being — James Bridle on More-Than-Human Intelligence

Turing and mirrors

Intelligence is a murky thing. While attributes like logic, reasoning, planning, problem-solving, self-awareness, and creativity are typically strewn about, scholars are yet to agree on a definition. I myself am partial to William Calvin’s view in How Brains Think—that intelligent systems are those that are able to guess well and find novelties under a broad variety of unclear situations. To Calvin, intelligence is more akin to jazz improvisation than a scripted concerto. Still, such definitions will continue to invite spirited debate.

In artificial intelligence, the polymath Alan Turing believed we could sidestep the question somewhat. He proposed an imitation game to analyze the conversational abilities of AI systems. If dialogues with AI were indistinguishable from that with humans, then it would be difficult for us to reject its thinking and intelligence, right? Turing’s test, of course, applies only to one facet of intellect. And it seems unfair to use humans as a benchmark since artificial life is unlikely to possess the lived experience that shapes our mode of speech. Even if the system was sentient, it may sound altogether alien.

As James Bridle writes in Ways of Being, intelligence comes in many forms. In this way, narrow tests for intelligence may tell us little about the minds we study. Some animals, for example, like the chimpanzee, bonobo, dolphin, or orca, will pass the ‘mirror test’—a popular hallmark of self-awareness and intelligence. Many gorillas, by contrast, routinely fail the mirror test and were previously assumed to be less intelligent for this reason. Scientists later learned, however, that gorillas actually dislike eye contact. They see it as a threat. So why should we expect them to behave like narcissistic humans? As Bridle notes, “approaches like the mirror test, which reproduce humans’ own obsession with faces… are simply inappropriate” or insufficient for many species.

Decentralized brains

Intelligence, Bridle reminds, is “more-than-human”. One illuminating example is the architecture of cephalopods like the octopus and cuttlefish. Octopi brains, in particular, are especially decentralized. They “extend throughout their bodies and into their limbs… Each of their arms contain bundles of neurons… allowing them to move about and react of their own accord, unfettered by central control”, says Bridle. They are “a confederation of intelligent parts.” This may help to explain their braininess in disparate domains. They have been observed to use tools, make plans, remember faces, build dens, and improvise under danger.

Some might view the social insects in a similar way. Here, we might think of each ant or termite as a feeble packet of neurons and signals. But in their sheer numbers, agglomerative algorithms, labor division, and chemical networks, elaborate structures and eusociality emerge. They include “modules” for a whole swath of complex behavior, from systematic search to backtracking to environmental learning and emergency-response rules. Today, variations of these distributed ‘brains’ thrive nearly everywhere on Earth. Ants possess a lineage that is more ancient than ours, and a history that may prove more enduring.

The wood wide web

Decentralized, non-human intelligence extends even to the warm beds of the forest floor. Trees and fungal threads known as mycelium, for example, network together to form the “nervous system of the forest.” The mycelium that grows on the roots of trees provide essential minerals like nitrogen and phosphorus to their compatriots. In exchange, the trees share with them the sugars they manufacture through photosynthesis.

What’s more, the trees use this fungi-root network to connect with each other. When “new seedlings emerge from the undergrowth, the mother tree infects them with the fungi, and uses it to supply them with the nutrients they need to grow”, Bridle notes. They have also been observed to use these networks to share nutrients with other trees and to warn of impending danger via chemical signals. All of this is a remarkable instance of the evolution of cooperation and risk-sharing between species.

This is no trifle matter. The wood-wide-web of information transfer and nutrient exchange is immense. “Globally, the total length of fungal mycelium in the top 10cm of soil is more than 450 quadrillion km: about half the width of our galaxy”, notes biologist Merlin Sheldrake. Indeed, the “forest [is] filled with a constant hum of unseen signs and unheard chatter”, writes Bridle. It is the sort of wheeling and dealing that should impress even the most self-regarding trader on Wall Street.

Sensory cripples

Unfortunately, our understanding of wider intelligences is clouded by the “twin hazards” of anthropocentrism and anthropomorphism, Bridle notes. We cannot help but see things through a narrow, blurry filter from our own frame of reference. It is easier to imbue non-human intelligences with human qualities than to see them for what they really are. In fact, as the biologist Edward Wilson observes in The Meaning of Human Existence, “the evolutionary innovations that made us dominant over the rest of life also left us [as] sensory cripples.” We seem to take immense pride in our taste for wine, ear for music, eye for art, and logic for chess. Yet there exists many natural and artificial systems that dwarf our raw ability in such domains. From the waggle dances of honey bees to the strategies formulated by chess engines, non-human systems exhibit bewildering novelty.

Simulation machines

Octopus brains, ant colonies, and forest nervous systems are reminders that our being and functions are embedded in interrelations and networks. Even our abstractions and algorithms, Bridle argues, move “in close concert” with nature. He points, for example, to the water-integrator, an analog computer that Vladimir Sergeevich Lukyanov built in 1936. To complete the construction of Soviet Union railways, Lukyanov needed to “model the thermal mass of materials” like concrete. Unfortunately, calculators at the time could not compute the necessary differential equations.

However, “Lukyanov realized that water flow was analogous to the distribution of heat and could act as a visual model of an invisible thermal process.” So the man built a “room-size machine made out of roofing iron, sheet metal and glass tubes”—using pipes, vessels, and pumps of water to undertake mathematical operations. The machine was a success, and Lukyanov and others soon realized its general and widespread utility. Future iterations found applications in geology, metallurgy, and missile ballistics. “Water computers continued to be used in Soviet institutions for large-scale modeling well into the 1980s.”

Moreover, Bridle uses the Lukyanov machine to say that “all computers are simulators.” They employ abstractions, algorithms, and models to describe and deduce various functions in our world. Not only that, he says that “the same is true of our own consciousness.” We are “simulation machines [that] become decision machines”, combining inputs with objectives, knowledge, beliefs, relations, context, constraints, and whatnot, in order to act. Unlike our machines and models, “we [sometimes] mistake our immediate perceptions for the world-as-it-is.” And here arises a complex feedback loop between the mind’s sky and our experienced reality.

Intelligent corporations

Again, the point that Bridle is making is that we live in a “more-than-human world.” For instance, when people ask him about the timeframe for true artificial intelligence, Bridle tells them that they are already here. They are corporations and institutions that run the world. Corporations, like intelligent computer systems, possess “goals, sensors, and effectors for reading and interacting with the world.” The difference being that corporations employ real human networks for sensing, planning, production, communication, operations, and any other function that aids in the system’s chances for survival and growth.

Similarly, in many respects, our collective selves are not dissimilar to the architecture of cephalopods and the combobulation of social insects. One wonders if some quasi-hive mind ought to emerge from this critical mass of human activity. It almost seems that way when bureaucracies and enterprises pursue self-preservation and self-perpetuation at all cost—neglecting the people they are meant to serve. But can an ant truly understand the colony that it is a part of? Humans are more intelligent, yes. But the systems, cultures, and institutions in which we participate are also many times more complex.

Darwin’s entangled bank

“The truth”, Bridle adds, “is always stranger, more lively and more expansive than anything we can compute.” This phenomena, he suggests, is well embodied in the fractal features of nature. In particular, when the mathematician Lewis Fry Richardson tried to obtain the lengths of nation borders, he was startled. He could not triangulate estimates that agreed closely with one another. But he then realized an even deeper insight. For many systems, the closer you look, the more contours and intricacies you’ll see. So how long is the border of the United Kingdom or Australia? Well that depends on how small a ruler you use.

Perhaps our sense of being and intelligence conforms informally to a sort of Richardson effect. The further we go, the muddier and more nebulous everything gets. At the very least, we can say that “intelligence is not a [reductive] collection of abstract modes”, Bridle writes, but “a stream” of qualities that is more than the sum of the parts. “Models of multiplicity”, he suggests, “are needed to make sense of this endlessly proliferating, teeming, oozing and entangling life”—a complex ensemble of process, connection, and interdependence. As the biologist Lynn Margulis writes in Microcosmos, “life did not take over the globe by combat, but by networking.” I think we can say the same about the architecture and diverse forms of thinking beings, and their dominance on Earth.

Ways of Being — James Bridle on Looking Beyond Human Intelligence