Nate’s Notes 13: The Singularity

Imagine a person living in the year 1700, who travels by foot and lives in a home heated by fire. Then imagine her traveling back in time 3,000 years. Life wouldn’t be too different. She wouldn’t be surprised to be living in a home heated by fire and to be traveling mostly by foot. But now imagine instead that she were to travel just 300 years into her future. Life would be unrecognizable. She would be shocked at the running water, heating and air-conditioning systems, flushing toilets, electrical outlets, radios, TVs, interstate highways, commercial farms, grocery stores, skyscrapers, jet airplanes, space shuttles, computers, cell phones, satellites, and the Internet.

Why does the first example of traveling 3000 years into the past differ so drastically from the 2nd example of traveling just 300 years into the future? The simple explanation is that technology has progressed at an exponential rate. For thousands of years, humans correctly expected their future to be much like their past. Exponential growth trends existed, but progress was so slow as to be imperceptible within one’s lifetime or even across many lifetimes. However, given our current technological capabilities, and current technological growth rates, our world may become virtually unrecognizable during our lifetime.

One of my favorite object lessons in my undergraduate business class is the paper folding example. It goes like this: Imagine a regular piece of paper: 8.5 x 11 inches. Now imagine that you fold that paper in half and in half, and in half, again, and again, and again. Then I ask my students, how many times would you have to fold the piece of paper until it reached the moon… which is approximately 240,000 miles away. Now I know you’re thinking you can only fold a piece of paper in half like eight times.  But use your imagination and imagine you could fold the piece of paper in half over and over again.

And actually to make it even more interesting, also make an estimate for how many times you would have to fold the paper until it reaches the sun, which is 93 million miles away.

For most of you I will tell you the answers and you won’t believe me. In fact, my students often tell me that they don’t believe the answers, even when I show them the calculations. But the calculations are simple.

Every time you fold the paper, you double the thickness, or in other words the paper gets thicker at an exponential rate. If you could fold a regular piece of paper in half just 42 times you’d already be to the moon. And if you could fold it just 9 more times, for a total of 51 folds, you’d be to the sun: 93,000,000 miles away. And as you can see, each time you fold the paper, the thickness just gets crazier and crazier: at just 66 folds, the paper is 1 light year thick, an incomprehensible 6 trillion miles.

But what does this paper folding example have to do with reality. Well computing technology has been progressing at an exponential rate since 1965. What does this sort of exponential improvement mean for technology products? Well the Apple iPhone is one such product—it has more computing power than all of NASA’s computers combined that helped send humans to the moon.

Other real-world examples illustrate the effects of exponential growth in technology. In 1990, biochemists set the goal to transcribe the human genome in just 15 years; yet, these same scientists had spent the prior year transcribing just one ten-thousandth of the human genome. At the then-current technological rates, the project would have taken ten thousand years to complete; however, due to exponential growth in computing speed (which the scientists accurately predicted), the human genome was decoded just 13 years later…two years ahead of schedule.

We know computers are improving rapidly and changing the world in the process, but there’s another reason this matters. Computing speed is the backbone of artificial intelligence, which means AI has been progressing at an exponential rate.

Just a few years ago, there was a popular business book published that opened with the following sentence: “With little effort, we can accomplish sophisticated tasks, such as recognizing faces or catching a ball, tasks that are far beyond the abilities of even the most powerful computers and sophisticated robots.”

That was just a few years ago. Since then, progress in AI “has completely exploded.  AIs now diagnose cancer, drive cars, trade stocks, translate languages, scan legal documents, write news articles, and write computer programs, and they do all of these tasks even better than humans in many cases. AIs even create music. In fact, the intro and outro music for this very podcast was created by AI.

But so what, why does this matter?

Well, if computing technology continues to progress at an exponential rate, AI capabilities will far surpass our expectations for it, in the same way that we underestimate the paper folding example. The future may not just be weirder than we imagine…it may be weirder than we can imagine. And it might be happening within our lifetime. In fact, simply extrapolating current technological growth rates out to 2045, yields an explosion of technology and intelligence that is unfathomable, an event called by some, the singularity.

And what if AIs eventually surpass human intelligence? In the same way that humans kill insects, often without a second thought, AIs might not be as benevolent to humans as we might hope. Plus, AIs would not need to surpass human intelligence by a wide margin to gain independence from humans. For example, chimpanzees share approximately 99% of human DNA. However, it’s that 1% that makes all the difference in terms of who sits at the top of the food chain.

Or maybe AIs will never surpass humans because the AIs and humans will merge. Computers continually get smaller and closer and closer to us. For example, few people ever leave home without their cell phones, apple watches, or ear buds anymore.

It is for these reasons that Stephen Hawking, one of the greatest scientists of our day, claimed that AI may be the biggest event in the history of our civilization, but it could also be the last. Others share Hawking’s concern. Elon Musk claimed that, “We need to be super careful with AI because it’s potentially more dangerous than nukes.” Bill Gates shares Musk’s concern saying quote “when people say [AI] is not a problem, then I really start to get to a point of disagreement. How can they not see what a huge challenge this is?”

Of course, if the progress of AI slows from exponential growth to linear growth, then we may not have much to be worried about. But it’s important to remember that each time we invent something new, we can use that new thing to invent additional new things.  Furthermore, in a world that is estimated to be 4.5 billion years old, we’ve become a so-called technologically advanced society in just the last few hundred years. What are the odds that we’ve maxed out technological innovation in such a relatively short time period. It seems very unlikely that progress will significantly slow.

In summary, be ready for big changes. And keep a close eye on artificial intelligence. If AI continues to progress exponentially, we will be prone to vastly underestimate it.

It’s a simple idea. Please take it seriously.