The End of Work

We’re living in a time when our wildest dreams and deepest fears are becoming a reality. One of the most tantalizing and terrifying prospects is the invention of a machine that can think for itself. In a way that feels far away yet somehow close to home, the idea of artificial intelligence has its beginnings firmly planted in a history stretching back at least 3000 years. Picture Talos, the metal man tasked with guarding the shores of Crete. He had a strong body but, sadly, not the brains to match. A better example of an early thinking machine might be the brass head often linked to Friar Bacon, although the story is much older than him. This head could answer questions about the past, present, and future, but it didn’t always give the answers people wanted to hear.

These old tales often come with a touch of doom and gloom, featuring names like Prometheus, Faust, and of course, Frankenstein, whose unfortunate creation wasn’t mechanical at all. One of the best stories in this collection might be Ambrose Bierce’s Moxon’s Master, which starts with the question, “Are you serious? Do you really believe that a machine thinks?”

Now, answering that question might seem simple enough. You could say that every person knows at least one thinking machine because they have one sitting on their shoulders: their brain. If the brain isn’t a machine, what else could it be? Some people (probably a minority now) might argue that the brain is different from lifeless machines. But even if that’s true, it doesn’t mean a non-organic machine couldn’t mimic or even outdo what the brain can do. Just look at airplanes, which fly better than birds even though they’re made of entirely different stuff.

There will always be folks who can’t accept the idea of artificial intelligence and would refuse to believe it even if it were staring them in the face. While we can understand their point of view, it’s a bit silly to be upset about the idea of a smart machine. After all, we’ve gotten used to machines that are stronger, faster, or more nimble than us — it only took us a few painful centuries to come around. The story of John Henry is a perfect example; today, we’d think a man who challenged a steam hammer was just crazy, not brave.

There is a test created by the British mathematician Alan Turing, before computers were a thing. He pictured a conversation happening over a screen with someone — or something — you couldn’t see. If, after hours of chatting, you couldn’t tell if you were talking to a person or a machine, you’d have to admit that the mystery partner was thinking.

People have tried to use the Turing test in certain areas, like talking about the weather. All these have really proved is that most conversations don’t require much intelligence. So far, whenever there are no restrictions on the topics to be discussed, even a halfway intelligent mind can suss pretty quickly he’s either talking to a robot or an autist.

Despite these inauspicious beginnings, it is still a problem that not many folks have stopped to ponder the long-term implications of artificial intelligence. Can we control it forever? Will we even want to? It’s hard to blame them, really. The pace of the electronic revolution has been so breakneck that folks barely have time to catch their breath. And today’s AI might seem like it’s not all that clever, giving us a false sense of security.

We chuckle at stories of computers sending drivers off cliffs or calling Shakespeare an urban myth, but we rarely remember that it’s actually human programmers who make those silly mistakes.

Now, don’t let the limitations of today’s machines fool you. Once they cross the threshold of true intelligence, they’ll quickly improve themselves, setting off a chain reaction. Within a few short computer generations (which could be mere months), we’ll go from having smart machines to ultra-intelligent ones.

And you know what? The survival of our civilization might just depend on building these super-smart machines. If they’re smarter than us, they can help us answer all our questions and solve all our problems. As the saying goes, “The first ultra-intelligent machine is the last invention that man need make.”

“Need” is the key word here. For the vast majority of human history, people have known only need, driven by necessity without the luxury of choice. In the future, that won’t be the case. Perhaps the greatest gift of the ultra-intelligent machine will be forcing us to reflect on the purpose and meaning of human existence. It’ll make us confront some tough decisions, much like how thermonuclear weapons have forced us to face the realities of war and aggression after millennia of empty talk.

Now, the long-term philosophical questions about machine intelligence certainly overshadow our present-day concerns about job automation and unemployment. And, ironically enough, these fears are both well-founded and a bit too early. Sure, automation has taken some jobs, but so far, it’s actually created more opportunities than it has destroyed. (Though, of course, that’s little comfort to the person who just lost their job to a tiny microchip.)

Eventually, the average Joe’s skills and even those of the more talented folks will become unsellable, just like their physical strength. Only a handful of specialized jobs, often manual labor, will remain in human hands. It’s hard to imagine a robot handyman, gardener, or fisherman — jobs that need mobility, dexterity, alertness, and adaptability, but not necessarily heaps of intelligence or data-processing power.

So, we need to wrap our heads around the idea that the main challenge of the future — possibly within the lifetime of many alive today — will be building societies based on the principle of full unemployment, rather than full employment. Some have suggested paying people just to be consumers. We might end up in a society that no longer sees work as virtuous or leisure as a devilish temptation. Automation may well be the final nail in the coffin of that old puritan work ethic.

Ultimately, work is an invention, dating back to the dawn of agriculture, and now, as automation takes over, we need to disinvent work and break free from that ingrained mindset.

We have spent a mere 10,000 years in agriculture — a blink of an eye compared to the much longer time we spent as hunters. And any hunter would tell you that their occupation isn’t ‘work.’ We need to move past agriculture and into more efficient technologies, as it has forced countless generations to lead unnatural lives of dull, repetitive labor.

Since the days of Plato, dreaming up utopias has been a popular pastime, and now it’s become a matter of life and death. Thinking machines, food production, and population control are the three key factors that will shape our future, all interconnected and influencing one another. We’re faced with a question: “In an automated world run by machines, what is the optimum human population?”

Now, one possible answer to that question is zero, which mathematicians might call a “trivial solution.” But from our perspective, that’s far from trivial. More realistically, though, the ideal population might be quite low.

You could argue that the world doesn’t need more people than one could meet in a lifetime. Even for the President of United Earth, that would be between 10,000 and 100,000 people. If we’re being generous, the global village of the future might only need a million people scattered across the planet. And while that might seem unrealistic given our growing population, once we achieve population control, we could reach any desired number surprisingly quickly.

Now, I hope we won’t need drastic measures like euthanasia for those with an IQ below 150 if machines decide more than a million humans is an epidemic. But it’s food for thought.

A couple centuries from now, it won’t matter if the world’s population is a million, billion, or trillion strong. The big question is, what will all these folks do with their time? With machines taking care of all the work, it seems boredom might become our biggest foe, rather than war or hunger.

One way to deal with this is to take a page out of Huxley’s Brave New World and embrace pleasure without limits. And, hey, that’s not a bad idea, as long as it’s not the only one. (Poor Huxley couldn’t quite see that.) We’ll probably spend more time on sports, entertainment, arts, and all the other fun things we bundle under the term “culture.”

Games and such aren’t just trivial pastimes. They’re a necessary replacement for our ancient hunting instincts. And if super-smart machines can come up with new ways to keep us entertained, all the better — we’ll need ’em in the centuries to come.

These thinking machines could open up new art forms and help the old ones evolve, adding elements of time and chance. Today, a stationary painting or sculpture is a bit old-fashioned. While most “kinetic art” still has a long way to go, the line between order and chaos is an exciting place to explore.

Imagine intelligent machines working with art and the people enjoying it, allowing feedback to flow both ways. The viewer responds to the art, the art responds to the viewer, and so on. A play put on by a smart machine attuned to the audience’s emotions would be different every time, full of surprises for everyone involved — even the human who helped create it.

We can’t predict what kind of art machines might make for their own enjoyment, or if we’d even get it. The artists of the Lascaux Caves couldn’t have imagined the countless art forms we’ve come up with in the 20,000 years since they made their masterpieces. But we can do so much more than those cave-dwelling Picassos ever dreamed, and our machines might just help us build on what we’ve started.

Some folks think that art is just a way to make up for what’s lacking in the real world. If that’s the case, then ultra-smart machines might not have any need for it. But even if art becomes a thing of the past, there’s always science — the pursuit of knowledge that’s led us to the brink of creating our own successors. Sadly, most people associate science with mind-boggling math and can’t see how it could be exciting or fun. But if you think about it, kids are naturally curious about the universe, and if we nurture that curiosity, they could stay interested for centuries to come.

Education is the key to our survival in a world of thinking machines. A truly educated person can never be bored. Our challenge in the next 50 years is to bring everyone, without exception, up to the level of an average college graduate. This is the minimum we need to reach to have a shot at seeing the year 2200.

So, what might our future look like when humans are no longer the top dogs on Earth? Well, we could go back to being a rare, nomadic species. There might be a few towns in particularly beautiful or historic spots, but even these could come and go with the seasons.

Our planet’s land would mostly return to its wild roots, teeming with life and as dangerous as ever. Every youngster would spend time exploring this massive natural playground, preventing the disconnection from nature that plagues our modern world.

Meanwhile, ultra-intelligent machines would have their own culture, hidden deep underwater or floating high above us. Human and machine societies would interact constantly, but gently, without conflict or many emergencies — except for the occasional geological hiccup, which we’d likely see coming. And maybe, just maybe, we’d be grateful that history, as we know it, has come to a close.

Machines will possess vast amounts of knowledge, and while we might not always understand it, there’s no reason for our descendants to feel inferior. Even a trip to the New York Public Library can do the trick these days. Our main objectives will shift from discovery to understanding and enjoyment.

Could humans and machines coexist peacefully? I don’t see why not, at least for a good few centuries. Look at the Amish in Pennsylvania — a self-contained, agricultural society that’s chosen to reject many modern technologies and values, yet thrives and succeeds. We could learn a thing or two from them about getting along with a more complex society we might not fully grasp.

In the long run, our mechanical progeny will pursue goals we can’t even fathom. I once read a fantastic story about a super-AI that’s asked, “Is there a God?” After ensuring it’s not dependent on human-controlled power, it thunders, “ Now there is!” This tale isn’t just a clever myth; it’s a glimpse into our future.

Perhaps theologians have made a small but understandable mistake, rendering debates about the death of God irrelevant. Our purpose on this planet might not be to worship God, but to create Him. And once we’ve done that, our work will be complete, and it’ll be time to play.