Whence and wither swarm intelligence?

by Thomas Moynihan.

“May not you and I be confluent in a higher consciousness, and confluently active there, tho we now know it not?”

—William James, 1909.

§

A year after the publication of Dune, in 1966, Frank Herbert released another novel, titled The Green Brain. Set on Earth in the near future, it imagines humankind waging all-out war upon the combined insect life of our world. With an armoury of pesticides and chemical bombs, the so-called International Ecological Organization aims to wipe out every last insect species. (Though there are plans to retain bees, genetically modified to be docile and subservient, which will fulfil critical environmental functions.) The entire planet has been split into two zones: civilized, bug-free areas on the one hand; on the other, dense jungles and uncultivated wildernesses, dominated by arthropods.

In response these human aggressions, inside a cave in Brazil, a brain develops. A “mass about four meters in diameter and half a meter deep”, filled with “alertness”, it coalesces from the collective action of a vast multi-species insect colony. It is maintained by countless drones crawling across its surface: cleaning, feeding, repairing, inspecting. Via a complex system of tactile and kinetic messaging, rippling messages constantly between its teeming subjects, the brain processes and relays information. It is “only vaguely like its human counterpart”: being not only “larger”, but also “more complex”.

Cogitating in it cave, as its colony crawls over it, it realises that humanity’s plan to eradicate Insecta will, of course, cause global ecological collapse and mass extinction. The brain devises a plan to prevent this from coming to pass…

§

The concept of “swarm intelligence” took hold in the late 1980s and early 1990s, in the context of robotics.

Prior to this, the term had already cropped up several times, stretching back as far as the 1920s and 30s, in relation to the wisdom of bee hives.

But it was in 1989 that engineers Gerardo Beni and Jing Wang presented a paper cementing the term’s present use. Here, they explored the ways a type of intelligence — a type of behavioural “order” — could emerge from the collective interactions of simpler, unintelligent robots.

Individually unempowered, when they come to together a far more potent agential force is produced from their coordination.

Thereafter, researchers began building artificial ants and reproducing, in silico, the flocking of birds. Swarm intelligence was an idea whose time had come. Today, it is a major frontier of artificial intelligence.

But, of course, as the above example from Herbert demonstrates, similar ideas pre-existed this 1990s efflorescence. Related notions — of “group minds”, “superorganisms”, and “hive minds” — hark very far back into the history of ideas.

§

Of late, we’ve become accustomed to claims that — through technological advancement, through the weaving of information across the surface of the globe — we may, ourselves, soon unify into one swarming mind. Sam Altman blogposts about what he calls the “merge”; transhumanists in the 1990s breathlessly claimed we are becoming “specialised cells in a macro-organic organism”. “We are Cosmic brain-cells”, they rhapsodized, building a planetary ganglion.

But the basic idea has surprisingly deep roots. Ever since ancient sages claimed our individual souls are but splinters of an extra-human mega-mind, ponderers have been equal parts allured and revolted by the suggestion. It became heresy in the Middle Ages; Hobbes called a similar beast “Leviathan”. What follows is a brief history of the idea of group minds, hive brains, and swarm intelligences.

§

The earliest intimations of such ideas probably trace back to Platonism and Neoplatonism, with their insistence our individual minds partake in one, unitary, mega-mind. Plotinus called it “τὸ Ἕν”. Writing during the 4th century BCE, Plato himself claimed that, like singular humans, cities themselves have “psyches”. Throughout his Republic, he assumed that these collectives have appetites, characters, and virtues, just like embodied individuals.

Later, in the 10th century CE, the Syriac thinker Yaḥyá ibn ʿAdī pronounced that, though we “are many in persons”, all “men in their true being are a single thing”: one encompassing “rational soul”.

Later, in the 12th century, the Arab philosopher Ibn Rushd (also known as Averroes) argued that our individual human minds are splinters of one, unified super-psyche. This thesis, which came to be known as “the Unity of the Intellect”, became controversial over ensuing centuries. Following critiques from Thomas Aquinas, the Catholic Church condemned it as heretical in the 1270s.

And yet the idea wouldn’t go away. The Renaissance scholar Marsilio Ficino likened Ibn Rushd’s imagined mega-mind to “an enormous monster consisting of many limbs and one head: a “compound made from man and from mind”, Ficino pictured the prospect as something octopoid and threatening.

Several generations afterward, in 1651, the English philosopher Thomas Hobbes saw a form of “artificiall life” coalescing from the ways humans interact not just with each other but also with their inventions and artefacts. Identifying the result as a type of synthetic mind, Hobbes described it as “an Artificiall Man”:

…though of greater stature and strength than the Naturall, for whose protection and defence it was intended; and in which, the Soveraignty is an Artificiall Soul, as giving life and motion to the whole body…

Governance provides the ligature of this titanic creature, Hobbes argued, whilst laws and economic exchange provide its “Nerves”.

Across ensuing generations, other thinkers returned to — and developed — this imagery. One of the most prominent was French sociologist Auguste Comte who, in the early 1800s, titled aggregated humankind “the Great Being”: a “social organism”, transcending the lives and capacities of all individuals to form an intergenerational, developmental ensemble creature.

In fact, Comte went so far as to claim that individuals don’t exist in their own right divorced from humanity as a teeming collective. “The disintegration of humanity into isolated individuals”, he declared, “is as vicious in sociology as is, in biology, the chemical decomposition of the organism into irreducible molecules, a separation which never takes place during life”.

§

As early as 1805, the German biologist Lorenz Oken — who believed animals themselves are made from countless cooperating mucous globules — noted queen bees resemble the ovaries of independent organisms.

They serve the specialized function of what would be an organ for an individual animal. Here we find the earliest roots of the idea of a superorganism.

Six decades later, in the 1860s, Oken’s countryman Johannes Mehring became the first to properly argue a beehive isn’t an aggregate, but one organism. A renowned beekeeper from the Rhineland, Mehring insisted “we must not imagine a swarm of bees as a herd of individual beings flying together, but that we are dealing here with a deeply interlocking animal organism”. He claimed that the workers, drones, queen, and hive together form one “a single organic being”.

The bees inside are a digestive system, he argued, whilst those that go roaming are like limbs. Mehring even compared the hive’s honeycomb to fat and body tissue.

Not long before, in 1851, the visionary German psychologist Gustav Fechner began making similar claims for the entire biosphere. Fechner claimed this unity possessed a form of gestalt intelligence more powerful than that which “lies in the capacity of all individuals”. Fechner proclaimed that, when combined, human brains form “the Earth’s brain”: pointing out this isn’t a literal frontal lobe embedded underneath the crust, but a synthesis formed from the “linking of our brains”. Concluding, he provoked:

…must everything supreme be a brain and be called a brain?

In Fechner’s eyes, summed together, our cognizing activities were generating a “soul” for the Earth. It wasn’t long before speculators were looking to the communications networks that were then crosshatching across the Earth: to the undersea cables, propagating telegraph signals across hemispheres. Already in 1871, German physician Rudolf Virchow was explicit: “telegraph cables” are the “nerves of humanity”, he announced.

They sensitize the globe, allowing it to communicate with itself, perhaps even to introspect, just as our own nervous systems grant us these faculties. Two years after, English philosopher Herbert Spencer explicitly related the spreading “electric-telegraph” to “cephalization”: the evolutionary process through which brains form and expand.

Indeed, Spencer had also already claimed that human collectives and societies are analogous to organisms, coining the word “super-organic evolution” to describe the type of powerful synergy produced by them. He described this as an “ever-accumulating, ever-complicating [production of] super-organic products, material and mental”, constituting “an immensely-voluminous, immensely-complicated, and immensely-powerful” whole.

From 1860 onward, Spencer had begun analogizing industry to digestion, commerce to vascular circulation, governments to nervous systems, and citizens to cells. Comparing the conglomeration of cities to the evolution of multicellularity itself, his conviction was civilization is something that is grown, not intentionally manufactured.

Spencer was far from alone: such ideas took hold in the closing decades of the 1800s. In 1892, American logician Charles Sanders Peirce ventured the following:

None of us can fully realize what the minds of corporations are, any more than one of my brain-cells can know what the whole brain is thinking. But the law of mind clearly points to the existence of such personalities…

In such an environment, wherein notions of distributed intelligence were taking hold, scientists even began suggesting extraterrestrial minds may well be organized according to this decentralized plan. The American geologist Alexander Winchell, in 1883, speculated there might be alien worlds housing highly-developed consciousnesses that are not “corporealized” in individual bodies as we are. They may be dispersed throughout “the abysses of the ocean”, he ventured, or perhaps even “plunged” into subterranean magma — whilst still retaining “consciousness and thought”. “It is conceivable”, Winchell mused.

§

As the 1800s shaded into the 1900s, more voices began arguing that social insect colonies should be considered not as aggregates but as singular problem-solving entities. In 1893, the influential German biologist August Weisman, in conversation with Herbert Spencer, presciently recognized that in respect to natural selection “the whole [hive] behaves as a single animal; the [hive] is selected, not the single individuals”.

But it was the American entomologist William Morton Wheeler who first used the term “superorganism” to describe the synergy of the swarm.

Wheeler noted that, as cells cooperate to forge an organism, organisms can themselves coordinate to render a whole more potent than its parts. Already in 1902, he was writing that the “ant-colony” is

…a complex of more or less heterogenous individuals, comparable to the Metazoan body, which is also a complex of units [or] more or less differentiated cells…

Wheeler wielded this insight to explain the previously unexplainable feats of ant-colonies: their industry, architecture, and behavioural complexity. They are essentially a brain, unrestrained from the confines of a skull, free to send tendrils of itself throughout its environment.

What’s more, Wheeler indicated that macroevolution itself can be understood as the gestation of increasingly powerful and complex wholes via the integration of formerly disparate, disaggregated individuals or parts. He wrote, in 1927, of the “irresistible tendency” of organic elements “to cohere and organize themselves into more and more complex emergent wholes”.

Knowing full well the amount of biological sacrifice involved in forming a superorganism, he recoiled, however, at the prospect of humanity tending further in the direction of the social insect.

Just as the mitochondria in our cells were once free-roaming organisms, so too did the ancestors of termites and ants cede various organs of independence and autonomy in submitting to the wider hive. Wings wither, eyes shrivel, even brains shrink. Unsurprisingly, it was precisely around this time that some writers began depicting the dehumanizing aspects of humanity’s own transition to something superorganismic.

Science fiction writers began imagining planetary hiveminds emerging, unintentionally, from global flows of exchange and thereafter coming to subordinate and enslave humans as mere nodes within their higher-order networks. Some celebrated this as the culmination of cerebral evolution and attainment of higher coordination; others foregrounded the loss of human autonomy and individuality.

Others were more neutral. The geneticist J.B.S. Haldane, in 1928, reasoned by analogy that, if the “co-operation of some thousands of millions of cells in our brain can produce our consciousness”, it becomes “vastly more plausible that the co-operation of humanity” may be produce something fathoms more potent. Continuing, he declared that just as “the unity of the body is not due to a soul super-added to the life of the cells”, so too may it turn out that producing the “super-human” may require nothing “external” to “human co-operation”. Should this ever take place, he pondered, let’s hope that it doesn’t override our “well-being” as individuals.

Elsewhere, still others, like Austrian-Irish physicist Erwin Schrödinger, began imploring — in 1925, in the wake of World War 1’s unprecedented destructiveness — that we must tend toward a superorganismic state, otherwise we will extinguish ourselves through war and bellicosity.

§

But undoubtedly the most prominent prophet of the coming world-brain was Pierre Teilhard de Chardin. Written in the 1930s, though published posthumously in 1955, his rhapsodic Le phénomène humain evangelized on our impending transition to something more unified.

For Teilhard de Chardin, the webbing and divaricating of information networks across the planet were providing the culmination of a deep evolutionary process that began with the sparking of sentience when nerves first began to weave and knot in the frontend of flatworms. Prophesying furthering oncoming waves of artificialization of the planet’s surface, and of biology and intelligence also, he claimed this would result in the inception of a planetary cogito. He described it thus:

We are faced with a harmonised collectivity of consciousnesses equivalent to a sort of super-consciousness. The idea is that of the earth not only becoming covered by myriads of grains of thought, but becoming enclosed in a single thinking enveloped so as to form, functionally, no more than a single vast grain of thought on the sidereal scale, the plurality of individual reflections grouping themselves together and reinforcing one another in the act of a single unanimous reflection.

This is how, in his words, intelligence “artificially perfects the thinking instrument itself”. He called it “mega-synthesis”.

Teilhard de Chardin even entertained the prospect of higher levels of cognitive synthesis: discussing whether populated planets may enter into communicate with each across sidereal distances. Then, he marveled:

Consciousness would thus finally construct itself by a synthesis of planetary units.

Others around the same time, were producing similarly mindboggling visions. In 1939, the English writer and philosopher Olaf Stapledon spoke of the cosmos’s “culmination” as the linking together of countless thinking planets into some “host”, some “cosmical society of worlds”. In his novels, he ecstatically described this unification as the “supreme moment of the cosmos”.

But, unlike de Chardin, Stapledon was far more sensitive to the potential negatives of this tendency. With reservation, he admitted:

Perhaps the spiritual perfection of the cosmos as a whole involves no such triumph of the enterprise of finite minds, but rather their partial defeat, much as the well-being of a living organism involves all sorts of internal, intra-organic conflicts, strains, and partial defeats. Of this we know nothing.

§

Teilhard de Chardin, a deeply religious and devout believer, dressed this up in an absolutely inescapable inevitability. He explicitly proclaimed that disaster or catastrophe absolutely could not derail the coalescence of his planetary soul. Due to his providential conviction, he claimed we “nothing whatever to fear” from the “cosmic catastrophe” or “biological disruptions” which may prevent our task.

However possible they may be in theory, we have higher reasons for being sure that they will not happen.

Today, such sentiment strikes as incredibly naïve. Teilhard de Chardin believed we “must reach the goal”. But history has no direction nor the future any guarantee. There is no inevitability. Nonetheless, the conviction — and disagreement about its normative valence — has lived on, flourishing particularly in around the millennium in the context of the early internet. Whether aware of the Teilhardism or not, in 2000 American investor George Gilder described “the web” as “a global efflorescence, a resonant sphere of light”: a ”chrysalis” giving birth to a deity composed from “billons of moving sparks from multimegahertz” chips and brains.

Some looked on this with ardent optimism, like, of course, Ray Kurzweil. Kurzweil talks of human and silicon minds merging not a “metaconnecton”, birthing one omnipresent, global cogito. Others presented more dehumanized and unsettling versions, such as the future depicted in Charles Stross’s 2005 novel Accelerando. Here, superintellects extinguish, then resurrect, humankind and the Solar System’s planets have all been rearranged into gigantic artificial “Matryoshka Brans”. There are “borganisms” and intelligent, uploaded consciousnesses of California spiny lobsters, which swarm over asteroids in the Mars-Jupiter belt. Stross sets the scene:

The solar system is thinking furiously at 1033 MIPS—thoughts bubble and swirl in the equivalent of a million billion augmented human minds. Saturn’s rings glow with waste heat. […] Small, crablike robots swarm the surface of Mercury, exuding a black slime of photovoltaic converters and the silvery threads of mass drivers. A glowing cloud of industrial nanomes forms a haze around the innermost planet as it slowly shrinks under the onslaught of copious solar power and determined mining robots.

Needless to say, what remains of humanity is no longer in history’s driving seat. Far, far from it: they are demoted insignificances in a world now incomprehensible to them. Stross describes:

From outside the Accelerated civilization, it isn't really possible to know what's going on inside. The problem is bandwidth: While it's possible to send data in and get data out, the sheer amount of computation going on in the virtual spaces of the Acceleration dwarfs any external observer. Inside that swarm, minds a trillion or more times as complex as humanity think thoughts as far beyond human imagination as a microprocessor is beyond a nematode worm. A million random human civilizations flourish in worldscapes tucked in the comer of this world-mind. Death is abolished, life is triumphant. […] Ecologies of thought are forming in a Cambrian explosion of ideas, for the solar system is finally rising to consciousness, and mind is no longer restricted to the mere kilotons of gray fatty meat harbored in fragile human skulls.

Stross’s Accelerando works because it produces painful ambivalence: it is deeply disturbing in its inhumanness but undeniably grandiose. Whether one thinks it presents a utopia or a dystopia — or something else entirely — will vary based on personal taste.

§

So, too, will answers personally vary to the query of whether it would be good, bad, or neutral if there might ever emerge — in the future, near or far, probable or improbable — some higher-order mega-synthesis. I ask myself the following (somewhat semi-serous) question: was it good for the mitochondria within our cells to be subsumed, ceding their autonomy in the process? From our present perspective, yes. But if any parahuman “mega-synthesis” or “meta-connection” took place, it’s difficult to see, in turn, much of what we could refer to as “our” current human perspective remaining on the other side. Do we really want to be braincells? In some ways we already are, but many extrapolative visions fill me more with horror than hope. In general, talking for the future is fraught and morally adjudicating outcomes even more so. Personally, I would not want any such mega-synthesis — if it’s even feasible — to happen soon. Most of my friends are human individuals, after all. Nor do I believe that the fact that something might have intensified in the past can be translated into a moral exhortation we must continue or intensify the trend. Indeed, the question of whether it even is an inherent fact is also deeply debatable. ‘Trends’ are deceptive; humanity’s track record for disinterestedly identifying them, abysmal. There’s plenty of fanaticism in this vicinity.

Regardless, the following words from Stapledon, written in 1937, remain compelling:

Immensity is not itself a good thing. A living man is worth more than a lifeless galaxy. But immensity has indirect importance through its facilitation of mental richness and diversity. Things are of course only large and small in relation to one another. To say that a cosmos is large is only to say that, in relation to it, some of its constituents are small. To say that its career is long is merely to say that many happenings are contained within it. But though the spatial and temporal immensity of a cosmos have no intrinsic merit, they are the ground for psychical luxuriance, which we value.

We still have little idea what mind is, let alone what it might one day be. It pays to be humble.

Thomas Moynihan is a historian of ideas. He writes about how worldviews transform over time.