It’s 2016. According to the movies, we were all supposed to have flying cars and hoverboards by now. It didn’t happen. In fact, if you were to transport Marty McFly from Hill Valley, 1985, to Silicon Valley, present day, he might be a little disappointed. Our world doesn’t look all that different from his. TVs are flatter, cars are rounder, but they still perform basically the same function. Airplanes and appliances haven’t changed that much. Toasters still only toast.
In fact, a time traveler from 30 years ago might be thoroughly unimpressed, until they came across one thing: our smartphones.
Thanks to some incredible advances in computing technology, we now carry around with us, all the time, devices that are much more powerful and more intelligent than even the most advanced computers from 30 years ago.
Moreover, we put those computers to good use. They help us navigate traffic, find out about new music, shop for new stuff, and show us things we might be interested in.
They’re able to do this not just because they have faster processors, but because they’re better at anticipating our needs and wants.
This is not easy for a computer to do. Because of the way they’re programmed, computers can do some things better than people, like processing numbers, analyzing data, and performing advanced mathematical calculations. But there are other things computers have a hard time doing that people do quite easily, like knowing what you’re going to do, after you say, “I have to be going now.”
This is because computers lack an inherent understanding of how the world works — they lack common sense. People develop common sense naturally. In their first few years of life, a baby will form, on average, 700 new neural connections every second. All those connections are moments where an infant is learning how the world works. As early as six months, babies can understand the casual implications of gravity. That is, they know that if a ball is tossed up, it’s going to come down.
A computer has no early childhood. If they are going to make judgements about the world, those judgements have to be programmed.
It sounds easy, but programming “common sense” into a computer is incredibly complex technical challenge. Computers have to be “trained.” Every misconception has to be adjusted. Every incorrect inference has to be corrected.
The scientific field dedicated to the solving this problem is known as “artificial intelligence,” or by its acronym, “AI.”
Whenever your smartphone suggests a place for lunch, or tells you about a band you might like, or gives you a heads-up that there might be traffic on your drive home, it is using AI to understand and anticipate your needs and wants.
Because of AI, we now have computers that can read X-rays better than radiologists, cars and trucks that can drive themselves, and even some robots that can improvise jazz.
But there are also deep-rooted myths and misconceptions about AI that I think are important to dispel. In Hollywood movies, artificial intelligence is often portrayed as some malevolent force, the hidden hand behind some robot apocalypse. Real AI is nothing like that. It helps computers make sense of the world, not plot its destruction. And as the science of AI advances, it will continue to improve industries from medicine, to construction, to academics, the arts, and even law.
That’s just a start. AI is a tool with incredible potential. Which is why we also have to prepare today’s students to work to understand and improve AI. America is already underinvested in math and science education. We cannot let an AI talent gap develop. Our high schools and colleges must increase investment in STEM courses, with a focus on the areas that are most valuable to the future of AI: statistics, data science, neuroscience, cognitive science, and even philosophy and ethics.
If we are going to solve humanity’s greatest challenges, from global warming to cancer, we are going to need the help of one of our greatest inventions: the computer.
And perhaps, with enough guidance, hard work, and investment in the science of AI, we might find ourselves, 30 years in the future, living in a world where the great promises of the future are finally fulfilled.
