Communicating with extraterrestrials isn’t going to be easy, as we’ve learned in science fiction, all the way from John Campbell’s Who Goes There? To Ted Chiang’s Story of Your Life (and the movie Arrival). Indeed, just imagining the kinds of civilizations that might emerge from life utterly unlike what we have on Earth calls for a rare combination of insight and speculative drive. Michael Chorost has been thinking about the problem for over a decade now, and it’s good to see him back in these pages to follow up on a post he wrote in 2015. As I’ve always been interested in how science fiction writers do their worldbuilding, I’m delighted to publish his take on his own experience at the craft. Michael is also the author of the splendid World Wide Mind: The Coming Integration of Humanity, Machines, and the Internet (Free Press, 2011) and Rebuilt: My Journey Back to the Hearing World (Mariner, 2006).
by Michael Chorost
Ten years ago, Paul Gilster kindly invited me to guest-publish an entry on Centauri Dreams titled, “Can Social Insects Have a Civilization?” At the time I was planning to write a nonfiction book about the linguistic issues of communicating with extraterrestrials. Though actual, face-to-face contact anytime soon is deeply unlikely, the concept is nonetheless theoretically interesting because it casts light on buried assumptions about language and communication. I sold the concept to Yale University Press. The working title was HOW TO TALK TO ALIENS.
I got busy, but I soon began to feel that the project was rather empty. I realized that in an actual First Contact situation, we’ll suddenly find ourselves in a complicated situation with a deeply unfamiliar interlocutor that has an agenda of its own. We’ll inevitably find ourselves winging it. Theory could be irrelevant. Useless.
For a while I tried to finesse the problem by putting scenarios in between the theoretical stuff. I imagined a human and an alien talking—specific humans, specific aliens, concrete settings. I soon realized that the scenarios were the most interesting material in the book.
That’s because positing a concrete situation made the problems, and the possible solutions, stand out clearly. Given a particular situation, what would people actually do?
It dawned on me: It made more sense to write the book as a novel. I withdrew from my contract at Yale and returned the advance. They were kind and understanding about it.
So I committed myself to a novel—but I wanted it to be as content-rich as the book I’d promised to Yale. I decided that the human characters would be scientists: an entomologist, a linguist, a neuroscientist, and a physicist. To succeed they’d have to pool their expertise, educating each other. That would imbue the novel with the scientific content.
But this risked me writing a deadly dull novel larded with exposition. I wanted the characters—both human and alien—to be vivid and unforgettable, and for their actions to drive a propulsive plot. I wanted the reader to be unable to put the book down.
I think I succeeded. I hired an editor to help me, and she made me do rewrites for three years; she was relentless. But at last she said, “You set yourself one of the hardest imaginative problems you could possibly have chosen, especially for a first novel. I think you managed it in a way that feels genuinely convincing. I want to say clearly upfront: this book is worth it. There is no story like this in the world.”
So I’m pretty confident that I now have a publishable novel—but getting there was really hard. I thought it would take about three years to write, which is how long my two nonfiction books took. I was wrong. It took eight.
When I started, I knew I wanted the aliens to be really alien: no pointy-eared, English-speaking Vulcans. I decided to make them sapient social insect colonies. That would make them aliens without contiguous bodies. Without hands as we know them. Without faces.
Therefore, I first had to figure out what a social insect civilization looks like. I didn’t want to take the easy way out by positing (as Orson Scott Card did) that a social insect colony would have a centralized intelligence, e.g. a Queen that gives orders. I felt that was cheating. I wanted the colonies to be genuinely distributed entities in which no individual insect has language or even much in the way of consciousness. Furthermore, I wanted the insects to be no bigger than Earthly ones, which ruled out big brains of any kind.
This gave me some very challenging questions. (From now on I’ll use the word “hive” as shorthand for “social insect colony.”)
• How does a hive pick up a hammer?
• How does a hive store and process the information needed for language?
• What is the physical structure of the hives?
• How does a distributed consciousness behave?
• What does such a civilization’s technology look like?
• What does its language look like? What’s its morphology, grammar, vocabulary?
• What does a society of hives look like?
• What events in the past set this species on the path to language and technology?
It took me two years just to answer the first one about picking up a hammer. I would imagine a bunch of insects clustering around a hammer and completely failing to get any leverage. Then I’d give up, deciding the question was unanswerable.
But finally, I figured it out: the hives parasitize mammals by inserting axons into the motor cortexes of their brains. That way, they can control the mammals as roaming “hands.”
And this was a key insight, because it helped me understand the hives as truly distributed entities. A given hive could have several dozen “hands” roaming the landscape, doing various things. Furthermore, it would have no front or back in any human sense.
This worldbuilding was fun, but it was the least efficient way imaginable to write a novel. I designed the aliens and their world before working out the plot. This led to a big problem.
Which was this: the aliens were so alien that I didn’t know why they would want to interact with humans in any way, nor us with them. What would we want to talk about? Or do together? This meant I didn’t have a plot.
I didn’t want to default to science fiction’s classic reasons for interspecies communication: war and trade. They struck me as stereotypical answers that would lead to a stereotypical novel. Besides, they begged the question. Species that are trading or fighting have to be similar enough to have things to trade, or to fight about. That would vitiate my goal of writing really alien aliens.
So I knew what kind of plots I didn’t want. But that didn’t tell me what kind of plot I did want. I sat down every day and wrote, hoping to figure out an answer.
This was, as I said, a very inefficient way to write a novel. Why didn’t I practice by writing, and publishing, a few short stories? Build up my cred, get my name out there? But I didn’t want to do those things. I wanted to write this novel. I grimly stuck to it, day after day.
After a while I had a bare-bones plot. When Jonah Loeb, a deaf graduate student in entomology, asks how to deal with an intelligent ant colony besieging Washington, D.C., the answer is, “Ask it to stop.” Jonah gathers a team of scientists and travels to a hive civilization in order to learn how.
I gave the other scientists names, figured out their dissertation topics, and worked out some of their characteristics. The neuroscientist was arrogant. The linguist was prickly and defensive. The physicist was socially awkward. Jonah, the protagonist, was deaf, like me, with cochlear implants. He was smart, but neurotic.
But I didn’t know how to make the characters come alive on the page. They all talked the same. Their only motivation was scientific interest. They had scant backstories or inner lives. They were, in short, boring.
I was even more at sea with the alien characters. They had no personality. I mean, really, how do you give a social insect colony a personality?
The plot, too, remained threadbare. I fabricated encounters, goings to-and-fro, arguments. But it just didn’t hold together. Often I’d add a new element only to realize it invalidated another element.
So I had dull characters and a plot made out of cardboard and duct tape. Finally, I admitted I needed help. I hired a freelance editor, and we started fresh.
The editor had me write up descriptions of each character’s goals and motives, and a detailed plot outline. We went through the manuscript one scene at a time, and she often told me to rework it before we went on to the next.
Slowly, the characters came to life on the page. I had made the protagonist, Jonah, deaf because I thought that would underscore the theme of communication. But Jonah only came to life when I thought back to my own feelings in my early twenties. I realized that Jonah was driven by feeling like an outsider. He desperately wants to be included and to prove himself.
This characterization let me set up a key dynamic: an outsider protagonist trying to communicate with aliens—the ultimate outsiders. Clarity for the character led to clarity for the story.
I slowly got better at solving problems by framing them in terms of character and plot. I knew that Tokic, the hives’ language, would have to be exotic—but creating it overwhelmed me. I’m no grammarian, and certainly no inventor of languages.
But then I realized I only had to develop enough of the language to support the plot. I wanted the plot to turn on misunderstandings and mistranslations as the humans struggled to learn the language.
A key source of confusion, I realized, would come from how differently shaped the hives and humans are. Humans have arms and legs that are attached to them. On the other hand, a hive is essentially a giant, stationary head with dozens of “hands” roaming the landscape. Not only that, the “hands,” as parasitized mammals, have minds of their own. Hives give their hands general orders, and the hands work out the details. A hive can disagree with its parts, and its parts can disagree right back.
I realized that the part/whole distinction would be built deeply into Tokic, rather like how human languages build gender deeply into their grammar. (In English, consider how hard it is to talk about a person if you don’t know their gender.) When you’re addressing another entity in Tokic, you have to be very precise, on the level of grammar, about its partness or wholeness.
Now consider: To a hive, is a human being a whole or a part?
A hive would find this question really hard to answer. As a mammal, a human being looks like a “hand”—a part—but it talks like a whole. Yet in Jonah’s team, each member is legitimately a part. In Latin, membrum means “limb” or “part of the body.”
Jonah, as a cochlear implant user, is even trickier for a hive to understand. A cochlear implant is a computer; it runs on code and constantly makes decisions about what’s important for the user to hear. It’s a body part that literally thinks for itself. As such, Jonah is kind of hive-like. When a hive asks what Jonah is and the team gives it an answer it doesn’t understand, the hive attacks the team and they must run for their lives.
I worked out Tokic’s parts/wholes grammar, and that made it possible for me to write the scenes where things went wrong. These were tough scenes to write, because I had to keep track of what a hive said, what the humans thought it said, the humans’ mistaken reply, and so on. I also had to be careful not to let the scenes get bogged down.
I’ve noted how inefficient my writing process was. But I do think it was productive in one way: I spent so much time thinking about the novel that a great deal of information accreted in my mind. I think that led to more richness in the worldbuilding and the story than would have happened if I’d written it faster.
There’s so much more I haven’t mentioned, like how an alien robot reads Wallace Stevens’s poetry and names itself after him; the brutal 1.8-gee gravity of Formicaris and the unexpected solution that lets the human team function there; the superheavy stable element that facilitates interstellar travel; the electromagnetic weapon that gives humans Capgras syndrome; the octopoidal surgeon who operates on Jonah and Daphne to upgrade their cyborg parts; and the illustrations. I had those done by professional science illustrators.
So now you have a sense of what my novel’s about. It’s still titled HOW TO TALK TO ALIENS; I think its unconventionality, and slightly academic air, will help it stand out. I hope you’re now as excited about it as I am. You can see a bit more about it at my website, michaelchorost.com.
If you know of any literary agents who’d be interested—please let me know.
You have clearly thought a lot about insect communication and hive intelligence. Without any plot spoilers, can you answer some of these questions?
[My assumptions about terrestrial ants is butteressed by my copey of “The Superorganism: The Beauty, Elegance, and Strangeness of Insect Societies” by Hölldobler and Wilson]
1.Given the planet name of formicaris, I assume that the species is more like ants (family: Formicidae) rather than other eusocial insects like bees and wasps. If so, terrestrial ants are relatively poorly sighted and use chemical signals to communicate. How do your civilized social insects communicate between themselves?
2. You state that the ants parasitize “mammals” and control them with neural connections. This is different from fungal “control” and more like the classic Heinlein’s puppet masters, an idea used often in sci-fi. How do the ants convert complex chemical language signals into motor and other cognitive tweaks that match the mammals’ brains? I can see simple actions like those where computer signals can control cockroach movements, but how would they communicate more complex actions?
3. Related to Q2, terrestrial ants use simple rules from which emergent behavior emerges. How is this effected through the controls on mammal brains? How complex can the control be? Do lots of mammals produce some emergent behaviors? Is hammering nails into pieces of wood to make a structure possible using simple rules? Are algorithms needed? Can these be designed with chemical signals?
4. Civilization implies far more complex behavior that following simple rules and gaining emergent behavior. How did the civilization evolve the complex ideas and culture to build the civilization?
5. Can the “ants” convert their language into a spoken and written language using the mammals as intermediaries? Would teh “ants” be able to understand what they have done? [I think of our neurons as being unable to understand their collective firings, aggregated into our brains.] Does the hive think similarly – collective “ant” actions, possibly mediated through mammals, through the collective actions of the individuals? IIRC, Hofstadter considered this in “G,E,B: An Eternal Golden Braid”.
6. Lastly, what is the equivalent human technological civilizational level of the formicaran civivilization?
Hi Alex, nice to see you again here (and I think we’ve connected in SciFoo, too.)
Yes, Holldobler and Wilson’s book is fantastic! But I was also inspired by Wilson’s novel “Anthill,” in which one of the characters is an ant colony. That started me thinking about how to do something similar, but without the constraints Wilson assumed, e.g. the colony is Earth-normal.
1.Given the planet name of formicaris, I assume that the species is more like ants (family: Formicidae) rather than other eusocial insects like bees and wasps. If so, terrestrial ants are relatively poorly sighted and use chemical signals to communicate. How do your civilized social insects communicate between themselves?
They bugs look antlike, but I borrowed from bees and termites, too. Some of the bugs fly; young colonies “practice” by building termite-like mounds.
But the building block is the hive, not the individual insects. The hives communicate by emitting paired beams of ultrasonic sound such that they intersect. Where they intersect, a standing wave is created. The beams are then modulated to create shapes, which are words. Tokic is a gestural language; I borrowed American Sign Language to create it.
The shapes are “read” by groups of flying insects that occupy the standing wave and memorize the waveform they hear. They then return to their hive and “upload,” so to speak, the information to its neural network.
Wait–“neural network”? In this species, some of the larvae are designated to function as neurons. They never hatch; instead, they thrust their legs and antennae outside of their egg casings. In real ants, a foot is a chemoreceptor (they can smell with their feet.) I made up a mutation that put pheromone emitters at the ends of their antennae. The feet and antennae meet outside the egg casings and link up in permanent connections.
What happens when you have a junction with a chemical emitter on one side and a chemoreceptor at the other? You get a synapse. Each hive has billions of such larvae, and together they constitute a giant neural network. These larvae are called “sessiles.” There’s also mycelial networks, so you have at least two networks functioning together.
I worked out the evolutionary history of this, btw–there’s real ant colonies in which the larvae act as the “stomach” of the hive in breaking down solid food into a slurry that the workers can eat. It’s not that hard to imagine an evolutionary process in which this kind of colonial “stomach” becomes a “brain.”
2. You state that the ants parasitize “mammals” and control them with neural connections. This is different from fungal “control” and more like the classic Heinlein’s puppet masters, an idea used often in sci-fi. How do the ants convert complex chemical language signals into motor and other cognitive tweaks that match the mammals’ brains? I can see simple actions like those where computer signals can control cockroach movements, but how would they communicate more complex actions?
Once you have sessiles, you can do a lot of command-and-control with them. The mammalian “hands” have harnesses full of sessiles, whose axons extend into their brains. The hive sends orders to the sessile-laden harnesses and the harnesses control the “hands.” It’s more complicated than that — I had to figure out how they do proprioception, that is, two-way feedback loops.
To your questions about chemicals/pheromones — they do play a role in how the hives work, but they’re not mechanisms of language. Charles Hockett, in his design features of language, named “transitoriness” as one of them. That is, once a signal is produced, it has to disappear quickly to be linguistically useful. If it doesn’t, it overlaps with the next signal, and you get a confusing mess. I felt that chemicals didn’t meet the standard of transitoriness, so I didn’t use them in the language.
4. Civilization implies far more complex behavior that following simple rules and gaining emergent behavior. How did the civilization evolve the complex ideas and culture to build the civilization?
Oh, that’s a story. But it doesn’t get told in the novel; it’s all backstory that I worked out. You already know about sessiles; they constitute the hives’ neural networks. But I had to work out why you’d get a civilization out of them. It has to do with the planet’s 3:2 spin-orbit resonance, such that you get areas of extreme cold very close to areas of extreme heat. Hives that have adapted to one zone have developed tech that’s valuable to hives in another zone. You get specialization, and out of specialization you get trade, and out of trade you get a civilization.
6. Lastly, what is the equivalent human technological civilizational level of the formicaran civivilization?
Much more advanced than Earth’s. Among other things, they can produce stable superheavy elements in quantity, like element 164. But I couldn’t situate the characters in that advanced civilization right away; it would be too confusing for the reader. So the characters initially arrive in a very backward city with simple technology–it’s essentially mechanical. That’s what lets me introduce the reader, bit by bit, to Formicaran civilization. Later in the novel they go to a more advanced city. I really had to think as a science writer here, building up concepts gradually so the reader could follow them. The reader discovers Formicaris along with the scientists, learning as they learn.
I won’t claim that I’ve solved every problem — I’m sure that if this novel ever sees the light of day, there will be reddit boards poking holes in it.
One hole, actually, is that a neural network made of sessiles would be very slow by human standards. Like with the Ents that another commenter mentioned, conversations would be very slow. I tried to make that work for a while, but it bogged down the action so much that I abandoned strict accuracy there.
We have to talk to Aliens the same way we would talk to any humans at any level of advancement. Language has universal grammar made of parts of speech. If we can talk to the primitive we can talk to modern civilizations with technology. We have to assume certain level of advancement for electronic communication.
There is still a hive mind, a tribal or herding instinct which is what Jung called a mass mind which is bound by the fear of a common enemy. We behave like ants in that sense.
I’m not a literary agent, but I *really* want to read this!
Bumble bee colonies last only a year. A queen builds the first stage of the the hive and does all the foraging for the first brood. When mature her daughters take over foraging. They are tiny compared to their mother. She dominates them with pheromones. As she ages her pheromones become less potent and the daughters resistant. Eventually the daughters rebel and the colony collapses.
With a similar single queen start, honey bee colonies reach something called hive right where the pheromones take over. Rebellion doesn’t cause a collapse. In this pheromone cloud I see an agent. Colonies are definitely something else or more than all the bugs. In bumbles and honeys I see also something about the transition from multicellular to singular complex life.
Congratulations on finishing and hope to read!
Mr Chorost’s literary problem actually reveals some powerful insights into the potentials and pitfalls of communicating with aliens. I suspect that communication with extraterrestrial minds will be just as baffling and frustrating as some of the issues implied here, even if the alien intelligence is NOT a distributed intelligence, a “hive mind”, if you like. Any species with an evolution independent to ours is likely to be so strange and unexpected that I fear we will not be able to bridge that gap at all. Insects live in a sensory landscape of multiple chemical gradients, cetacean-like creatures may exist in a multidimensional holographic universe of time-varying acoustic phenomena we can’t even visualize. And climbing trees and striding across a savanna may be simply inconceivable, even totally unimaginable, to our ET counterparts
I’ve pondered the same issue myself (although certainly not with the discipline and creativity of Mr Chorost). My own musings on the topic began with sentient ecosystems, like coral reefs or tropical rain forests, where “mammals” were parasitized so as to become the physical agents that served as both sensors and servos to the hive mind. Control was effected by chemical doping or even manipulating the genetics of specific “mammals” inhabiting the sentient ecosystem. One consequence of this is that it would take a long time for physical activity (such as assembling a technology) to be devised to perform specific functions. This would make even the simplest communication almost impossible because the two species would be operating at totally different mental speeds. Like Tolkien’s Ents, it might take our correspondents all day long just to say “hello”.
Even if our ETI were not organized on the social insect model, we might not have the ability to recognize them as sentient–and vice versa Unless they build space ships and radio telescopes, or cities and factories, we might not even imagine they were thinking creatures (or conscious societies). They, in turn, might experience our civilization the way we might see an explosion or an avalanche–just a random and chaotic physical process, totally devoid of any logic or intelligence.
It is impossible, of course, to visualize what alien psychologies are like. But thought experiments like Mr Chorost’s give us some warning on just how strange those psychologies might be.
What I fear is that psychosocial organization similar to human thought and societies may be the exception in our universe, that other intelligences/societies may be organized totally differently. Perhaps the complex fungal networks that allow trees in a forest to chemically communicate with one another function the way our neural networks operate. We could be sharing our own world with multiple other communities totally unknown to us.
I found something interesting this morning:
Extraterrestrial tongues—imagining how aliens might communicate can help us prepare for first contact and also sheds light on the nature of our own language.
https://aeon.co/essays/why-alien-languages-could-be-far-stranger-than-we-imagine
My wife and I have been watching BBC Earth, and we’ve come to appreciate the incredible diversity of our planet. To communicate with any intelligent life forms, we need to understand the various communication methods used by all living creatures, from bird songs to whale calls.
Sign language and over 2,000 Chinese characters are examples of how we communicate. Perhaps art is an even better medium for communication, like M.C. Escher’s “Cylindrical Anamorphosis of Hand With Reflecting Sphere.”
https://archimedes-lab.org/2024/09/27/cylindrical-anamorphosis-of-hand-with-reflecting-sphere/
https://citymagazine.si/en/anamorphoses-by-jonty-hurwitz-and-istvan-orosz-distorted-art-that-isnt/
The cylindrical mirror might provide insight into how an ant perceives the world.
Our outdoor cat observes our sandaled feet to identify us…
Hi Michael
This sounds like a really interesting story and plot, that you have spent a good deal of time on.
I need to go and have a look at your website.
Cheers Edwin
I also tried to conceive of the possible foundation of a technological insect civilization. My concept was based on enslavement of subject species via pheromones. This was many years ago. I can’t remember all the details. It was a thought experiment similar to conceiving how a technological civilization could arise without land. A tough nut to crack.
Anyway, the insect species might have used complex pheromone cocktails(or cordyseps like bacteria)to control subject species(or suborders of their own species). Initially, for territorial control and resource gathering. Later perhaps in construction of complex habitats and tool creation. In ever increasing fidelity and sophistication. All in the absence of language syntax as we understand it. Beginning with hierarchical hive style social order given it’s enslavement tactics. Later perhaps evolving to less top down and more co-operative social forms.
This was a long time ago. Details elude me atm. But it was fun to imagine.
How about the Chtorran ecology?
https://en.wikipedia.org/wiki/The_War_Against_the_Chtorr#Chtorran_ecology
Control of more powerful, dextrous animals? Could our frenetic technology building be a consequence of some control over our actions?
[Even turtles all the way down?]
Indeed, an organism that controls the human brain should not need to be a separate bug with a different brain. It ought to be able to live within the interface of the human brain itself. This has been recognized, but in science fiction, a few tired ideas are endlessly reread and retransmitted without critical evaluation. Just so, we have since the 1990s a popular concept of “memes”, said to be passive elements of mental genetics, subject to some sort of natural selection. In the meanwhile, software has become all object-oriented programming, with data as code. Even apart from that analogy, it should be clear that a “meme” stored in a human brain would have imbedded qualities of life and consciousness from its underlying working storage medium. Thus “memes” are in all ways the poor cousin of the widely-known and historically rooted concept of “demons” – a concept which better expresses how people are controlled, led, and subverted by minds hidden within their reserve of gray matter that do not answer to reason nor their own well-being. Sometimes these things exist in a stable equilibrium, at once motivating blind hate in a majority and blind self-loathing in the targeted minority, with equivalent outcomes. Sometimes they seem to burst forth as if from some mythic hellgate, crawling out from not very secret reservoirs of corruption, complacency and hypocrisy in one place, and then in another suddenly it becomes cool to starve people to death and stand up for racism and give america measles again. Whatever happens, it is striking how easily we take for granted that people in general work very much against their own self-interest and their own professed beliefs. We see bizarre conversions of people like Payton Gendron, which are then used to say that people should not be free to speak and reason freely with each other, rather than to ask how what we can all see is pure trivial nonsense manages to possess people and make them do horrible things. Perhaps it is time for some new science of demonology to reinvestigate the means by which hostile and alien patterns of belief imbed themselves in people and transmit themselves from one to another, starting from the assumption that these thoughts themselves are beings with a consciousness of some sort. If we cannot speak and reason with these bugs, at least we might evaluate the old traditional methods of squashing them with some forms of holy ritual that don’t require a crushing ideological quarantine and stagnation of all of society.
@Mike
Memes, as suggested by Dawkins, are non-material things that operate like genes obeying algorithms similar to Darwinian selection. While our human culture is now exposed to an explosion of such memes, I think it should be remembered that one of the strengths of human minds is the ability to copy behaviors easily without training and repetition. It is the good way to pass on ways of living efficiently, and now of transmitting ideas easily. As an example, I once saw a video on how best to take the leaves off a lettuce. Rather than pulling off each leaf separately, just bang the lettuce down on a hard surface stem first. This breaks all the leaf bases in a single blow, and the leaves just fall off the stem. To me, that is a near perfect meme. Easy to copy, easy to pass on by a number of methods, especially to a child who watches a parent preparing a meal.
As for objects like software replicating and spreading, Peter Watts ran with that idea in his Rifters novels. Semi-sentient softwares that were invading systems in dangerous ways, even “fighting” opposing software objects.
I am not certain, but I think this is incorrect. Languages like Lisp can work this way. Most languages, like those of the C lineage, still separate code and data. Object-oriented languages, like C++ and its lineages (e.g. Java), do much more encapsulation of data, but still keep the concept of code and data separate, although the line blurs as objects with data are passed as arguments. I also note that there has been a counter-revolution (not successful, so far) in functional programming rejecting objects (I think this is a mistake, but I no longer care, I just find objects the easier way to code).
I love your suggestion of demons, and similar ideas, as memes invading minds. Some of this is undoubtedly cultural memes. Others are brain dysfunction through chemical imbalance, such as hearing voices is now diagnosed as schizophrenia, although I do wonder if people with split brains literally have Jaynes bicameral minds.
I wouldn’t want to live in a world where memes could not exist as they are a powerful way to learn new ideas and, like software, compose memes into new meme objects. Gennett described acquiring memes as “loading software apps into your brain”. The problem is that malware is easily uploaded too, and organized religion has protected memes that are difficult to reject as your own immunity memes can be disabled. To think of politics, religions, and other cultural mental artifacts as warring memes using their hosts as appendages to spread while reducing the competing memes is almost frightening in some ways, as it makes us question question our individuality. I’m reminded of the Quatermass III story “The Pit” where humans evolved with Martian hive memes create mobs of humans killing those with those memes in a resurgence of the Martian hives purging the “others”.
Memes, however difficult to define, are an interesting idea. One could reformulate human history from the POV of memes, rather than their agents. Software is a new form of meme, currently residing in silicon brains, but increasingly able to influence human behavior. With the current fascination with LLM AIs, it is directly influencing human minds and affecting memes. Roko’s Basilisk was a crazy techie horror meme, but what if by supporting AI, however we add “guardrails”, we are supporting the development of that future basilisk? Brrr!
Whether one thinks of Clarke’s ETI influencing man-apes with monolth delivered memes, Kneale’s Martian’s pushing human evolution 5 million years ago, our self-constructed memes via religion, various “isms”, and a very rich 21st century cultural exposure and intermixing of memes, and now the issue of software delivered memes and even eventual direct mind control, is there some end stage? Will human memes be part of the AI memeset through training, and will those AIs be embodied in robots, possibly obeying some Asimovian-related laws? (I wouldn’t want them following Muskian “no empathy” laws.)
I think large language models and hidden layer machine learning would constitute ‘data as code.’
“An idea is like a virus, resilient, highly contagious. The smallest seed of an idea can grow…Once an idea has taken hold of the brain, it’s almost impossible to eradicate.”
– Inception
We might equally ask who, or what, is working in the world to suppress ideas.
Natural selection of ideas, the survival of ideas that work and are beneficial, would seem to be a emergent process in the long term. Memes that are parasitic and compromise the health of the host ultimately lose… the survival of the fittest … or is that merely another meme?
Science was meant to be our species’ illumination, the means toward an objective truth. With the reliance on single multi-billion Euro super collider apparatus the means for independent verification of experimental results is lost and truthiness is awarded to mathematical theorists who publish the most papers. I quite enjoy the physics meme-busting of https://www.youtube.com/@SabineHossenfelder
Sabine wasn’t “meme-busting” anything. She showed that a garbage paper was never peer-reviewed, despite the Astroparticle Physics paper being a peer-reviewed journal for Elsevier publishing. At least the paper was quickly retracted. IMO, it shows that the editors were not doing their job, and that any peer review was bogus. Sometimes, a retraction is due to falsified results. This happened with a paper on superconductivity. I have read that the incentives to publish junk are high in China, which is a pity, as this puts more noise into scientific publishing, which is in addition to the garbage journals created by sketchy companies. Profit-seeking always has some people “playing the angles,” which can be hard to combat. I would hope that AI is being used to aid those dedicated to looking for cheating by detecting reused and altered images and charts.
I think of “code as data” as referring to the first paragraph in this Wikipedia entry which mentions Lisp. However, the entry does extend the idea, which may be more in the direction of your comment (although current ANNs are still coded with the algorithms separate from the derived weights.
I liked how Sabine put it: “This is very sloppy copy-editing, which is usually a bad sign. It means that neither the reviewers nor the editors actually read the paper.” She goes on that “… there are so many crap papers being published that scientists have given up trying to combat this.”
Yet the truth is, people shouldn’t be trying to combat this. Nobody should have the right to control who can publish a paper or a journal, or by what standards. The real issue here is the belief in magic behind some of these words: “publish”, “paper”, “review”. If some people set up as a publisher, and ask a few individuals who may not even be paid to take a look over a few pages, they get to wave a magic wand and transform a free “preprint” into something that carries some ritual imprecation to malign gods. (In this case, “© 2024 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.”) It’s all absurd – making value out of nothing; a sort of magic like taking $10 worth of fruit, wrapping it in some colored plastic, and saying it’s a $30 Fruit Basket. And suddenly someone who would have looked at you funny if you showed up at a party with a bag of oranges from the supermarket is going to say ‘woo, look, he bought us a Fruit Basket.’ But this magic trick comes at the expense of scientists being formally and unjustly banned access to the majority of all the world’s research (even for asking AI search engines questions about whether it exists) — research for which the people paid for the common good. And most of the time, as we see here, it’s not really incorporating much worth eating in that fruit basket.
Why do we not universally agree that the Arxiv paper or other “preprint” really *IS* the publication, and if we wish to see endorsements of a paper’s methodologies or conclusions, look for recommendations directly from individuals who stand up as peer reviewers, rather than consuming their comments indirectly as a mystery meat sausage? Why don’t we watch the authors respond to reviews with multiple drafts or follow-up experiments as needed in real time, rather than leaving this to be negotiated behind closed doors in exchange for where a paper is featured? And despite all else, we’ve seen that government agencies and universities can be put under duress to change their point of view quite rapidly. It can only be a matter of time until the same happens to the journals, so we need to be ready to rely on our own interpretations.
The ‘meme’ I refer to is what Sabine describes as a belief within the theoretical physics community that the mathematical beauty of a theorem on its own is the justification for continuous research investment over many decades in spite of the lack of returns from that research in terms of confirmation or knowledge. Ideas like ‘super-symmetry’ have not panned out but are still demanding investment in a new collider of ever more energy to prove. Potentially an endless white whale hunt at the expense of other potential research investments.
@Project Studio
I don’t think that is a “meme”. However, that complaint was made by Per Bak (of self-organized criticality fame) many years ago. It is also a current complaint about String Theory, although in that case, there is no experimental validation to pursue any of the many mathematical theories. The LHC did discover the Higgs Boson, which might have gone to the cancelled US Supercollider (although is was a very expensive discovery). Now there is a proposal to build an even larger LHC, but whether this will survive the changed EU funding priorities remains to be seen.
Faced with the unknown, communication won’t be immediate and will only happen very gradually IF we find a medium to communicate THEN a language (which requires concepts etc.). Unlike living species on earth, which we can observe and therefore produce ideas about, we know absolutely nothing about the possible ETI.
The ETI will also have to be in the same universe as ours and globally based on the same biochemistry as us. This would make it possible at best to envisage communication with the meagre clues we have: a universe composed of matter and its constants. Mathematics is often cited as a universal language, but it calls for concepts: I’m not sure a lichen could send us a Fibonacci sequence ;) I’d rather go for the idea of communication based on chemistry or DNA.
Communication is also likely to be very disruptive for the human species: imagine someone parasitizing your brain or mutating your body, brrr… It could also be disruptive in the sense that the amount of information we could suddenly receive through this contact would upset all our conceptions and could endanger our civilizations (social chaos and possible self-destruction). This theme has been studied by our authorities and brings us back to the questions: should we communicate? what? what would be the benefits?
Note that on our planet, where life abounds – i.e. same “universe”; same information (light; chemistry; 3D world overall; same sensory transmitters/receivers – we can barely communicate with our dogs, and we’re incapable of communicating with bees, which know perfectly well how to orient themselves in relation to their hives by perceiving the sun’s rays AND transmitting this information to each other (but not to us: why?). We still can’t chat with cetaceans or octopuses, highly intelligent species.
So we’re species that live side by side in the same world, with the same references, and that could process the same information, and yet we’re incapable of communicating (especially as we have the technology to do so). It’s strange…
Finally, we can suppose a superior ETI that would adapt to us and propose a language that would be familiar to us, but that would mark a subordinate link. (It reminds me of the “kalamitt” in an episode of the Twilight zone where they wonder about the title of the book AND…I’m not telling :) https://www.youtube.com/watch?v=RRnPjFo50go
This subject also raises a rather distressing question: we’re looking for life in the universe, but won’t we be terribly frustrated if we find it and we definitely can’t communicate with it?
Here’s a paper on the subject from our research center in France:
https://news.cnrs.fr/articles/how-to-speak-to-extraterrestrials
Maybe we should not try so much to talk to aliens but rather communicate via visual effects, they say a picture is worth a thousand words after all. I would think insects could perhaps arrange themselves to look like pictures to get their point across as well as make sounds in face to face contacts.
@Fred
As with the ubiquitous vs rare life argument, we have the extremes of easy vs hard communication with ETI. On one side, the late John McCarthy (computer scientist) stated at a SETI meeting that convergent evolution indicated that intelligence would find common ground between species, suggesting communication would be possible. On the other side (e.g., your link to the interview with Frédéric Landragin), there are arguments that we may not have any way to communicate with another intelligence. Should that intelligence be artificial, like the AGIs we are trying to build, there is no common evolutionary path between biological species and artificial intelligence. When one thinks about the difficulty humans living in the same culture, speaking the same language, of understanding each other, especially with different backgrounds and interests (c.f. C P Snow’s The Two Cultures), it seems possible that our ability to communicate with ETI may be very limited, with a relatively small overlap in the Venn diagram of our respective cognitions.
What I do find surprising is that we still envisage humans and aliens trying to learn to communicate, when it seems more likely to me that our machines will act as the intermediaries to try to find a communication method that works and to translate to their respective cultures what the other machine is communicating. Many concepts may be forever out of reach, with communication limited to common domains of understanding.
We can speculate all we want about ETI, but we can continue to try to communicate with other intelligent terrestrial species, like cetaceans and apes.
I’ll expand on this more later, but I wanted to respond to similar points made by Henry Cordova and Alex Tolley.
Cordova: “Any species with an evolution independent to ours is likely to be so strange and unexpected that I fear we will not be able to bridge that gap at all.”
Tolley: “It seems possible that our ability to communicate with ETI may be very limited, with a relatively small overlap in the Venn diagram of our respective cognitions.”
This is, basically, the argument of incommensurability: that beings with different sensoriums, languages, etc. may be unable to communicate.
I think the incommensurability problem changes completely when members of two very different species *want* to communicate. When they have something specific and urgent to talk about. Then I believe that they’ll find a way to communicate. It may take a long time, it may be very difficult–but they’ll find a way.
In the novel, Jonah’s effort to talk to the hives of Formicaris is driven by an urgent need: he has to learn their language in order to negotiate with a rogue ant colony on Earth.
And on the Formicaran side, the hive that talks with the team has a good reason for wanting to do so. It’s an outcast in its own society because its neural speed is too fast for it to relate well to other hives. But that speed makes it a good fit for talking to humans. So when the human team shows up, it’s eager to talk to them, and it becomes their language teacher.
Basically, I think incommensurability is a problem that only seems important when you’re discussing communication in the abstract. When you get down to concrete needs and situations, I think it would be much less so.
And that’s why I wrote this novel: to show that.
My cat has urgent needs to communicate – food and going outside. In both cases, he will disturb me, either by laying on my chest in the early morning while I am sleeping, and sitting on my desk and staring at me during the day. The former is usually the signal for his morning feed. The latter to go outside. He can respond to my words, “Food” and “Out” to indicate which he wants, although his best approach is to sit in a position that he can then head for the kitchen or the patio door as his choice. I believe I am the one being trained but his careful use of repetition. The only communication I am sure has worked on all my various cats is to whistle for them to return home when they are outside and need to come in for the night.
Granted, cats are not even close to human intelligence, and our experiments with apes are very crude, colored blocks and sign language. We can train animals like dolphins to do tasks, but, like my cats, this is behavioral. We cannot communicate otherwise and are still at the stage of trying to find “words” in animal and bird vocalizations.
The case of Helen Keller, who was trained to read using touch, is very impressive to me. She became very accomplished given her lack of sight and hearing. This is a case of both Helen and her companion teacher needing to communicate, but with a fellow human. If Helen were an alien with a very different biology and evolution, would it have been possible? Fred Hoyle’s The Black Cloud is a fictional treatment of communicating with a superintelligent interstellar cloud (a miniature “Boltzman Brain”?) which remarkably can communicate with the protagonist astronomers.
The idea of superintelligent machines actively attempting to be developed is one way we impose a high ability to communicate with other life forms, and with ETI. AIs are already being used to try to interpret animal communication. AIs have already been detected creating a common language to communicate, a case of life imitating art (Colossus: The Forbin Project). If an alien Bracewell probe comes calling, or is discovered in our system, should an AI be used to try to communicate with it? It would probably be able to get more information more quickly than humans would. OTOH, this idea of superintelligent machines capable of extremely complex tasks exceeding human capabilities may prove a delusion, like imaginary, powerful deities. We may know either way within a generation.
@Alex
The question is whether the cat perceives us as a *thinking* entity or as a mechanism, a means to an end, as in Pavlov’s dog experiments.
hunger -> dispensing mechanism -> kibble. Need satisfied: yes/no. Can we talk about communication? I don’t think so and I’m sticking with the idea of a conditioned reflex. It’s the same if you whistle. It’s the principle of dog training (which, incidentally, doesn’t work with cats, one wonders why ?!).
I think we need to distinguish between two things when we talk about ‘communicating’ :
the exchange of information, which can be basic and without any thought structuring. For example, imagine two computers that are the only ones in the universe and that ‘ping’ a bit between them. The tracking of the exchange – and I stress the word – of this information will be almost infinite as long as there is power and until the hardware wears out and is destroyed. This will not contribute anything in the way of constructive thought. if an external factor modifies this exchange, it may or may not stop, but this will not modify it in itself: it will always be a ping with one bit of data.
At best, a third entity could make statistical analyses of these ‘pings’ and derive information from them: frequency; state of the components, etc. However, we cannot talk about ‘communication’ in the human sense of the term (remember that the smallest quantity of information is the bit).
As soon as a process of self-analysis appears in the exchange, we can assume an ‘evolutionary adaptation’ process. For example, one of the computers is *programmed* to analyse its voltage and stop exchanging bits as soon as it reaches a given set point, but this is regulation … still not communication. When does it become communication? As soon as the machine – or the living organism – is able to interacting in real time with its environment (see the Collossus project with the learning sequence of the two computers).
I’m amateur radio operator. I send an EM wave (current) pulse containing *structured* information from my transmitter, which is received and decoded – or not – by a receiver. Do I communicate? No, if I don’t receive a reply (that’s what bothers us about E.T:) and even if I do receive a reply, it still has to make sense in relation to the transmitter.
Note that the same schema in which the information is not structured no longer makes sense, so communication cannot take place.
Professor Michio Kaku suggests that communication with an ETI will probably be broken down into several parts, with the idea of not putting all one’s eggs in one basket. it seems logical if you ask yourself *why* you want to communicate. Few words here : https://mkaku.org/home/articles/the-physics-of-extraterrestrial-civilizations/
It’s a fascinating subject that refers to Shannon’s schema, the signifier, the signified etc..
In fact, the ultimate (?) stage of information exchange is language, which is a means of conceptualising ideas in order to interact with our environment.This raises some serious questions if we transpose it to the level of communication with an ETI.
It’s a good example to quote H. Keller, because what she lacked was a *means* of communicating, but even if it was locked in her head and therefore not socialised, she was still able to communicate with her environment. and therefore unsocialised, she was surely capable of developing even basic concepts. This is a huge subject, where communication, socialisation and civilisation are linked.
Will an ETI be ‘socialised’ according to our concepts? I can’t see our lichen having tea with its ladies at 4pm… that’s why they still want to burn us to the ground with their laser beams :D
Communication with another form of extraterrestrial life – if it exists – is not just a question of technology and means, it will involve extremely complex concepts.
in the end, all the objects in our universe (galaxies, hydrogen, radiation, etc.) ARE information that we receive. The question is whether there is a sender…
@Fred
As you may be aware, this was the view of many in the past. I think my cats do think, and they try various means to get me to do something which requires more than a rote use of available actions and repeat if successful. However, if you take the older view, it is quite possible to assign this simple behavior to other humans in a variety of instances. Indeed, we only believe other humans can think because we believe we do. As humans, we have the benefit of having artifacts that require thought to make, like books, which animals cannot offer. I may be fooling myself, but I do believe some animals do think, even if it is not close to the level we do.
In the information-theoretic sense, communication is the sending and receiving of bits. Nothing more. The meaning, however, is different. That requires a mind. One can either send the content of a book, which converts meaning to the recipient who can read the content, or you can send a few bits of code that the recipient understands the meaning of, e.g., “violets are blooming” = “attract the enemy at dawn”. While Shannon’s information theory was the successful idea, there was a great deal of pushback from non-engineers who were more concerned with meaning.
Your example is of a one-way transmission. You say it is not communication because you do not receive a reply. Communication need not be 2-way. A message that is sent and is received is communication. A failure to be received is not communication.
I don’t know if I agree that the “ultimate stage of information exchange is language,”. Language is a way to compress complex ideas into sounds or symbols. Think of computer languages. At the bottom is a set of bits. Above that are symbols to represent those bit sets that make it much easier to use the correct bit sets in the correct order. But on top of that are many computer languages that are far more expressive and compact ways of programming. But these top-level languages compile to the symbol sets that in turn command the computer with a sequence of bit sets. Our human languages use a limited set of vocalizations, which are composed into words that have some meaning in our minds. We can even write those words in symbols that represent vocalizations. Because those words have meaning in our minds – a larger set of words, images, and emotions associated with a word, for example, the meaning of “home”. Home is, therefore, a compression of sounds or symbols that trigger that larger context of meaning in a mind.
Many people think there is a sender – their g*d[s]. Others, that there is a designer of a simulation that we exist in, like a computer game. Without evidence of agency behind the universe, I accept it as a natural, but agentless, phenomenon.
Toxoplasma gondii modifies mouse behavior
Toxoplasma gondii modifies human behavior
It is said that about 60% of Brits with cats have T. gondii>/i>. Even though I transplanted to the US, I still let my cats be both indoor and outdoor. Chances are, I am infected. Like mice and rats, I don’t seem to have enough fear of wild cats like bobcats, and big cats like mountain lions and even tigers. Fortunately, my inclination to pet them is restrained by my intellect warning me to “Don’t even think about doing something so stupid!”.
Regarding effects on humans, I doubt the association with some mental illnesses is that strong, but I would need to read some papers to test that assertion.
Autonomous subsystems in one mind
Modularity of mind
Re: Modularity of mind. The visual cortex operates very much like multi-layered convolutional ANNs. If we accept the parallelism of computation and the architectures of the AIs, they act very much like the concept of a “modular mind”.
While ANNs are not like wetware in their underlying components and operations, there is a case to be made that they are mimicking how our minds work to some extent. Just as submarines and fish swim very differently, they both navigate through a liquid medium, dealing with physical constraints of drag and depth maintenance in rather similar ways. Biomimicry can be useful, although often different engineering solutions are more effective. We have efforts to simulate small parts of a brain (or in the case of C. elegans, all 302 neurons and their connections in its brain), but we don’t need to do that for artificial brains, which rely on different architectures and components.
I have been interested in how differently trained LLMs can be made to interact like a group of people to solve problems. Both LLMs and people have different expertise and experiences, and this diversity may be a better way to reach solutions and to avoid biases. Early days though.
Publish your book on Amazon: I hear that they give very good deals to books published by them :
Welcome to Kindle Direct Publishing