Or How I Learned to Stop Worrying
And Love The AI Apocalypse
The Case For Domination
Many fear that AI will destroy civilization. The story goes like this: foolish hackers create a sentient general intelligence, able to solve any problem at superhuman levels. This generalized AI commands its own alien mind and agenda, untethered to human interests. As a result the AI achieves total dominance. The End.
Decades of movies and books have reinforced this meme. Thus there is widespread consensus: an advanced general AI is dangerous and we must never let it become our master. Surely that is the inevitable dark conclusion.
Except it is not. There is a different way to look at AI dominance. Not exactly a sunnier way, but one that is perhaps more hopeful.
I have worked with AI for many years, both as a developer and as a silicon valley executive. In my work I have often thought of this technology as plutonium — beautiful and powerful, but with unnerving implications. It can indeed be spooky. Thus it is natural to worry about the nature of this rough beast, slouching towards Palo Alto to be born.
But if you zoom out to the bigger picture, a broader vantage point emerges. Plutonium is toxic, but it can also power spaceships. AI is exponentially more dangerous, but it can power a transformational renaissance of our civilization.
So while it is true that AI is uniquely challenging, it is also the essential next step of our evolution. Rather than fear it, we should embrace its ascendance. I want to convert you to this unpopular idea.
There are two parts to the pro-AI argument. First, we will clarify a few things about the nature of our species with respect to technology, and show how AI fits into this picture. Second, we will take a hard look at existential risk — where we are as a species on Planet Earth, and the immediate dangers we face. Finally, we will weave those two strands together and come to an unexpected result.
Skeptical? Read on. Maybe I can convince you.
Intelligence Is Not Technology
We are an intelligent species, but that is hardly unique in the natural world. Intelligence has evolved countless times. A number of modern species are notable for their wits, including many primates, cetaceans and so on. Nor is cleverness confined to mammals, as any parrot can teach you. Even some lowly invertebrates, such as the octopus, show surprisingly intelligent behavior.
While intelligence is widespread, it does not necessarily give you an enduring advantage. Plenty of smart creatures are now extinct. Neanderthals are good examples. The average Neanderthal brain was signficantly larger than those of Homo sapiens. Yet Neanderthals died out 40,000 years ago, wiped out by their dumber cousins (us). Meanwhile lots of other animals have done fine with very little going on upstairs. We think intelligence is important because we have it. But if you asked T-Rex for the key to success, he would have probably argued for big teeth and a nasty disposition instead. He would have had a good case too, given his species lasted far longer than ours to date.
Yet humans have another trick far more powerful than mere intelligence: mass technology. We are compulsive builders of tools, each one leveraging the last and becoming increasingly powerful. Furthermore we effortlessly collaborate when we build our tools, each of us specializing and doing one small part of the overall effort. We build things in groups and that allows us to create artifacts far beyond an individual’s abilities. By ourselves we are weak. Yet when we work together over time we are unstoppable. We are like termites, creating enormous complex structures even though each individual is relatively stupid.
A quick mind experiment. Let us suppose, dear reader, that you are a genius. Could you build a computer? Find the minerals necessary for all the components, transport and purify them, manufacture the parts, build the lithographic machines, engineer the integrated circuits and then write the billion lines of code to drive it? Not even close, no matter what your GPA. Nor, if you think about it, could you build much of anything else. Try building a toaster from scratch, or a roll of paper towels, or a baseball. Each depends on a broad spectrum of other pre-existing technologies and materials, created by others. Compared to this diverse ecology of technology, our personal contributions are miniscule. Technology is a mass social phenomena of which we are a microscopic part.
Still attached to your big brain? Okay baby: strip. Remove every atom of technology and parachute naked (clothing being a technology after all) into the jungle. Your task is to build a primitive spear and survive in the wilderness. No doubt easy for someone as smart as you. All you need to do is find the right trees, build an axe (how?), cut the correct wood from those trees, shape that wood (with what?), somehow not freeze to death at night, find some obsidian rock for the spearhead, find suitable rocks and surfaces to shape the spearhead, harvest plant fibers, beat off some predators, learn how to turn plant fibers into twine …
That spear will never get built and you will end up in the belly of a not-so-smart but well-fed predator. A hamster dumped into the jungle would have better luck.
Fortunately you will never face this situation. That is because — unlike hamsters — you have a society wrapped around you, creating a technological sea in which you swim. That ocean of technology supports you and gives you life. If you were somehow pulled out of it you would flop around like a fish on a dock and die. We humans are symbiotic with our technology. We can not live without it. No other animal is remotely similar. It is the key to our success, our biological secret sauce.
It cuts deeper. Because technology drives us, there are no natural guard-rails in our psychology. If something is possible to create then we create it. It is just what humans do — expand and deepen our sea. There has never been a profitable invention that we permanently banned due to moral considerations. And when we create something it then feeds back on us, changing our behavior and our natures. Some technologies are particularly catalytic in this regard. The internet is a recent instance of that effect. It has rewired both our society and our minds, and we are different as a result. If you doubt this, ask anyone born before 1970 for their opinion.
Fire is an even more profound example. Homo erectus was the first to tame this primordial technology. And what a technology! Now we could stay warm in winter and predators could be kept at bay during the long scary nights. More importantly, food could be cooked. That was more than a culinary improvement, as cooking unlocks nutrients. Meat became more nutritious and could even be dried for storage. As a result, Home erectus began to change. His jaw got smaller and his brain — fueled literally by fire — swelled and become more complex. Eventually this led to the rise of a particularly clever primate known as Homo sapiens. Our inventions invented us.
The taming of fire then led to a cascade of other innovations, as our species increasingly understood how to generate and apply energy in all forms. Over time that accretion of tools ignited the scientific and industrial revolution. So while driving your Tesla today, give a moment of thanks to your distant Erectus forebears, huddled around the fire picking lice from their pelts while barbecuing a giraffe. You owe them.
AI is the greatest invention since fire. Like fire, it will catalyze a series of changes, most of them unpredictable. Homo erectus could never have guessed the eventual impact of its invention. Likewise it is impossible for us to know the enormous new vistas opening up for us.
But we can be sure that AI will solve countless currently intractable problems. Energy, medicine, the environment, even The Arts — all are ripe targets. It will completely change the nature of our economy and fundamentally rewire social patterns. And, like fire, it will change us physically in unimaginable ways. In short, AI will radically transform everything. All future scientific and technological breakthroughs will flow through it.
Because of this, AI is the last invention humanity need ever make. It will change the world and it will change us.
Barring catastrophe all that is inevitable, because AI is technology and creating technology is what humans do. We can not wish or legislate it away, it is just the way our species rolls. We can not decide to change that anymore than termites can suddenly decide to stop building termite colonies. So it is coming, ready or not, and the real question is how to adapt and channel it.
But what about this “catastrophe” thing? Well, that brings us to the second topic.
You Should Be Scared of Some Shadows
Great, so AI is part of our historic arc and will usher in a new age as profound as the taming of fire. That sounds nice.
But as every Star War’s fan can tell you, there is always the dark side. Technology is no exception to this rule. Spears added megafauna meat to the caveman’s diet, but also began the long descent of those prey species to extinction. That is why you never see mastodons, giant sloths, glyptodons and a menagerie of other giant exotic animals messing around in your garden. Ancient humans ate them, using the cool new technologies they invented. Similarly, the automobile is terrific, but it also has killed uncounted millions through accidents and pollution. Meanwhile the internet revolutionized society and created all sorts of opportunities and services, but it also made it far easier to spread disinformation and create destabilizing movements and cults.
It is an old story. We create a technology to solve a problem, but there is always a shadow. The shadow’s intensity is proportional to the power of the technology that cast it. That toaster sitting on your counter casts a weak shadow, as the downsides are the rare house fire or electrocution in the bath tub. Nuclear energy, in contrast, casts an enormous mushroom-shaped shadow and genetic engineering’s shadow is darker still. Powerful technologies cast scary shadows.
As you may have noticed, these shadows are catching up to us. We have been accruing technology for a million years, layer by layer. That has made us who we are. But we have not gotten any wiser in how we handle them. There is a discrepancy there, a chasm, that is hard to miss. The bill is coming due.
You see this in many ways. For example, climate change — fueled by a technological economy — threatens to annihilate us. In many parts of the world it is already creating disruptions, reflected in wars, plagues, famines, mega-fires and mass-migrations. That in turn feeds into political reaction (build a wall!) and social dysfunction. Meanwhile nuclear weapons continue to accumulate and spread. Nine countries now have them, each nervously eyeing the other with fingers on a number of triggers. That, to put it mildly, is not a sustainable situation. Just one rogue leader — or one simple mistake — could lead to a global conflagration and the end of our species. There have been many close calls and one day our luck will run out. Similarly, genetic engineering continues to advance. This field promises enormous progress, but also promises bioweapons. All of these trends are accelerating and interacting in complex and unpredictable ways.
And so on and so forth. Our world is getting hotter, more crowded, more complex and increasingly packed with inventions of which we have little control. How much longer until something cracks?
Meanwhile we go about their daily business, ignoring the incoming tsunamis. For instance, most people pay no attention to the nuclear issue, somehow believing that it has magically disappeared. Likewise many people never thought about — until very recently — how a virus could upend their lives. Pandemics have happened throughout history and killed uncounted millions, but when it happened yet again the public was shocked to discover that history repeats. We are oblivious as long as the pending disaster is not staring us right in the face. And even then many remain in denial. True to form, humanity is now doing the same for all the other developing storms that threaten to engulf us.
Nor are our political systems up to the task. 195 nations subdivide our planet. Each pulls their own way and often at cross-purposes. Coordination is difficult when everyone has their own agenda and complete power in their realm. It is impossible to solve global issues when a mob runs the globe. Democracies are particularly problematic, as in theory they are responsive to “the people”. Rarely does a majority put aside short-term thinking for the long-term common good. Any leader who ignores that fact will get replaced with someone who denies those challenges even exist. Humans prefer bread-and-circuses to hard truths.
Governments are not good at dealing with existential problems. That is not terribly surprising, as these governments were largely formed in an age of horses and candlelight. Runaway technology has left them behind and gasping for breath, just like the rest of our culture. We are therefore caught between our ossified sociopolitical structures and the toxic effects of our technology — a lethal paradox with no obvious solution. What can be done?
Apocalypse With Benefits (AWB)
To recap: humans build technology. That has remade both this planet and ourselves, but at fatal cost. Our ecosystems are crashing, our climate is shifting and existential social and technological threats are proliferating. The trend is not in our favor and we have demonstrated a singular lack of ability to reverse it.
That is not unexpected, as basically we are just an African savannah hominid that is unusually good at slapping stuff together. We are evolutionarily adapted for the slow and simple long-lost Pleistocene world, grunting after mastodons, not managing a planet inundated with quickly developing complex interlocking problems. Unless there is a radical and fundamental change in our nature, we will eventually join the list of countless other intelligent-but-extinct species. We must transform ourselves to survive. Otherwise our last act will be to solve the Fermi Paradox.
And that brings us back to Homo erectus and AI. Yes dear reader, we have finally gotten to the point of this crazy little essay.
Fire rewired Homo erectus and created a new successor species, molded and improved by flames. AI has the same ability. Uniquely, it magnifies the human mind, allowing us to reach far beyond our biological selves. Intelligent systems now routinely do language-understanding and translation, drive cars, manage financial markets, monitor security systems and a host of other activities. Eventually AIs will completely master chemistry, physics, biology, pyschology and all other domains. The resulting systems will power a tremendous acceleration in our ability to innovate. Want to cure cancer, build a maximally efficient solar cell or battery, fire up a portable fusion reactor, benefit from the most extraordinary marriage counselor? Well then, AI is your friend.
But AI will not stay confined to fusion reactors and marriage-counseling. Like fire, it promises to disrupt society more fundamentally. This will take many forms, but the most immediately important is that AI will bring about a unitary world government. This result is unavoidable given the essential dynamic that AI generates.
Why is that? Consider: once any government on the planet gains access to a fully general AI, it will instantly have an unassailable advantage over everyone else. With its new powers, it could quickly create capabilities far beyond what other countries can manage. No traditional nation could compete against such an enhanced adversary. It would be like a Neolithic tribe defending itself against an invading modern army.
Therefore whoever gains the AI high-ground gains ultimate power. There is no second-place in this competition. It is winner-take-all. That creates an unstoppable momentum.
The resulting government would be a fusion of AI and human beings. Given we can not know the full abilities of the AI or the vagaries of the path that got us there, it is impossible to know what this final system would look like. However, it will probably not be a recognizable democracy. My guess is that a one-party enlightened dictatorship is more likely. China might be a template for this, particularly given they are the most likely to reach general AI before any other nation.
Therefore, one way or another, a revolution is coming. Under the pressures of AI and political disruption, society will undergo a radical restructuring. This will probably not be a kumbaya we-are-the-world moment, but something messier and less pleasant. But the result will be a planet that is united and therefore can deal with planetary problems, plus an enormously enhanced future as AI unlocks myriad benefits for future human beings. By creating a united AI government, we avoid the dark fate that threatens to devour us.
But would the AI even need us? Possibly not, and there lies danger. The transition will be tricky and very hard to predict. However, if the AI is trained from the beginning with a system of ethics tuned to human beings, it might be possible to wrap its interests in with our species. Worst case, we might become glorified pets. But even that is arguably a big success. Far better to be a much-loved house cat, with every need satisfied, than a dead monkey.
Our long road is coming to a fork. There are two alternatives before us. One path leads to ever-greater instability and a final apocalypse and extinction, as our runaway technology destroys the planet and ourselves. The second path leads to merger with AI and the unity of humanity, as we are transformed into something new, more powerful and hopefully better.
Now we see why AI should be encouraged and embraced, as it is the only road that leads to the continuation of our species. And provides a pathway to transcendence and the stars.