86-Billion Neurons are Not Enough
Restaurant menus display body parts of animals that we regard as less intelligent than we are. At a recent dinner in a fish restaurant in Boston, I had buyer’s remorse watching the octopus arm served to us as an appetizer. Octopuses have a complex nervous system and excellent sight, and are among the most intelligent animals.
They have a brain and a nervous system containing 500 million neurons, two thirds of which are in the nerve cords of the arms. In comparison, the human brain is estimated to contain a hundred trillion connections among 86-billion neurons. As the server approached our table with the dish, I realized that this must be the reason why the restaurant chefs felt privileged enough to serve octopus parts as an appetizer. Surely, they would feel bad serving an entity more intelligent than us for dinner.
The first terrestrial entity to contest our place at the top of the food chain would be artificial intelligence (AI) systems. Gladly, their body is made of silicon chips so it cannot digest human body parts. But it could digest the human mind. One of the biggest existential challenges to humans is mental health in the age of AI.
Once the number of connections in AI systems exceeds a hundred trillion — the corresponding value in the human brain, humans might be manipulated and controlled by AI systems just as octopuses are manipulated on their way to restaurants. Within the past decade, social media served junk food to human minds and resulted in a polarized society, full of hate towards people we never met. Adding AI to social media without caution could lead to even more polarization and hate.
86-billion neurons are not enough. We tend to hate people we do not know rather than converse with them. One is left to wonder how peaceful would human history be if the human brain had a trillion neurons and could not be manipulated so easily. This question can be studied in the distant future by developing AI systems with trillion connections.
That technological feat might require a new architecture that goes beyond silicon chips. Future technologies might take inspiration from nature, which started on Earth from a soup of chemicals and ended up over 4.2 billion years in a human brain — which consumes only 20 watts instead of the gigawatts needed to power our current AI systems.
How would natural selection look like in a world of super-human intelligence? Does “survival of the fittest” — the principle advocated by Charles Darwin, imply that the survivors have the highest intelligence?
Not necessarily so. The Nazi officers in concentration camps during World War II were brutally efficient but not as intelligent as many of the 6 million Jews they executed. Who will dominate our future history?
There is a scientific path to answering this question. Our telescopes offer us an opportunity to conduct a census in our cosmic neighborhood. By surveying billions of habitable exoplanets throughout the Milky-Way galaxy or by encountering extraterrestrial technological artifacts near Earth in the form of space trash or functional devices, we can get a statistical sense of the hierarchy in the cosmic food chain. Who is controlling whom, either literally — on a physical dinner plate, or metaphorically — through mental control?
To find out, we need to change priorities within the mainstream of astronomy and invest billions of dollars also in the search for extraterrestrial intelligence, rather than focus on the search for microbes. Seeking superhuman intelligence or extraterrestrial brains with more than 86 billion neurons — either artificial or natural, will be a sign of scientific humility and a marker of our own intelligence. However, as of now the 2020 Decadal Survey of Astronomy and Astrophysics sidelined this goal and defined the task of searching for primitive forms of life as its highest priority. This is not the best way to recognize smarter partners on our cosmic block.
Given the fact that astronomers did not recommend allocating billions of dollars to the search for superhuman intelligence, we are left to speculate about whether “the smartest survived” there. The brilliant manager of the Galileo Project in search for extraterrestrial technological artifacts near Earth, Zhenya Shmeleva, who also specialized in biology, alerted me to the novel titled `The Invincible’ by the Polish writer Stanisław Lem where alien machine lifeforms evolve through relentless environmental pressure and selective annihilation on the exoplanet Regis III.
Remnants of earlier military machines, likely from a lost civilization, undergo evolution, where survival, rather than intelligence, drives complexity. The most advanced system of intelligent machines fails under its own cognitive load and high demand for energy, while simpler units of microscopic insect-like robots proliferate as adaptive, decentralized swarms.
When faced with a threat, the minuscule robots assemble into huge “clouds” that move with a high speed and can incapacitate any intelligent threat by being brutally efficient but yet entirely unintelligent at the individual level, like the Nazis during World War II. In a modern military world driven by drones and a technology world where AI is shaped by optimization and performance metrics, Lem warns us that superficial survival instincts without intelligence and empathy, and power without intention, are plausible outcomes of natural selection among alien civilizations.
As in any blind date in our personal life, we must listen rather than speak in our first encounter with aliens. Instead of prioritizing microbes from the bottom of the terrestrial food chain, astronomers must contemplate in the 2030 Decadal Survey of Astronomy and Astrophysics the possibility that our date partner, sitting on the opposite side of the interstellar dinner table, is smarter than us. Our long-term survival might depend on that sense of humility.
