When you pair artificial intelligence applications with quantum computing, you come up with something truly extraordinary
By Eric Ladizinsky
For years, we’ve heard big stories about the promise of artificial intelligence (AI). Many scientists and futurists have waved their hands in the air, promising that AI will have a big implication in many fields, from factory floor automation to stock trading and medical decision-making. But so far we’ve seen very little in the way of real commercial products. Innovations like Siri and Google Voice are nice, but not world-shaking. Eric Ladizinsky argues that a new AI era is about to arrive, brought on by the application of quantum computing theory to AI applications. After decades of broken promises, we’ve seen big advances in AI in recent years, brought on by ever-improving machine intelligence. It is no longer the stuff of science fiction. Instead, AI could reinvent our world, and what it is to be human.
When people talk about the future of artificial intelligence (AI), they often speak in colossal terms. We hear exciting predictions about self-driving cars that effortlessly and safely get us to work. Or we learn about traveling humanoid robots that do everything from factory work to emergency rescues. These enchanting tales of the future are often couched in the larger story of constantly evolving conscious Internet — we users represent the neurons, our texting and emails serving as nerve endings, and the electromagnetic waves zipping through the air are neural pathways.
Sometimes, however, the predictions about AI are not so rosy, and may even be downright apocalyptic. We’ve all seen movies about robots that become sentient and conniving creatures, identifying humans as an enemy that needs to be destroyed. Yes, I’m looking at you, HAL, from “2001: A Space Odyssey.”
As a quantum physicist, I often think big — both about artificial intelligence and the future of computing — and I’m firmly optimistic about the future. Why? AI is essentially any machine that mimics a living and intelligent organism. One day, AI-based devices could help detect and obliterate the first signs of cancer. Or they could pull together the disparate pieces that help us save the biosphere from global warming. I also envision far more prosaic applications for artificial intelligence — ones that will impact every customer-facing business around, from car dealerships to hospitals to judicial systems.
Put simply, I see a world where machines are actually more human than we are.
Let’s think this through for a moment, using one of our least-favorite ways of spending time: dialing into a company’s customer service line. Imagine engaging with a service or sales rep who always knows exactly what you want or need, whether you’re prone to compromise or argument, or even if you’re tired or happy. Sound desirable?
Let’s consider something more profound. We also know that doctors, when presented with patients with identical symptoms, don’t always reach the same diagnosis and treatment plan.
Research shows that different doctors, presented with the same symptoms in the same patient will come up with a different diagnosis and treatment plan. We also know that different judges, when presented with the same set of facts and circumstances, don’t rule identically. Incredibly, judges who rule one way in a court case are capable of issuing the opposite ruling when given the same set of facts and laws several months later.
There’s a simple reason for this. Humans aren’t always rational. We’ve got our own ideas, agendas and judgments — all of which influence how we behave and interact. That’s fine when you’re dealing with mundane decisions about what movie to see or meal to eat, but it would be nice to imagine a higher order of intelligence applied to something as critical as our nation’s laws or our personal health.
With artificial intelligence, machines will instantly recognize a person, detect her mood and tap into vast databases to arrive at the best solution for her particular concern. The technology to do this is incredibly complex, and it’s currently enormously expensive and error-prone. But that will change, as these things always do. When it does, robots won’t be annihilating the human race. They will help us solve some of the world’s most intractable and commonplace problems.
Technology’s limit: Humans
Recently, I set out to buy a new car. I headed to a local dealership, where I met a nice salesman. I was not just a customer looking for new wheels — to him I was also another commission. I knew the model of the car I wanted and I also knew that I was in no mood to haggle or settle for something less. But when the model I picked didn’t come in the color I wanted, the salesman tried to talk me out of the color and I wanted and into the color he had. I wound up spending a few hours on the phone, calling local body shops to find out what a paint job or a high-end vinyl wrap would cost. The experience was time-consuming and, ultimately, really annoying.
My experience was not unique. I suspect it happens thousands of times every day. Customers (or perhaps potential customers) are consistently stymied by the agendas, biases or egos of others when transacting business. My attempt to buy a car ran into the same underlying problem: the salesman had his own preconceived ideas about what would happen when I walked in the door. I know that the tone of my voice changed and my eyes narrowed as I grew increasingly frustrated with the salesman. But all he was thinking about was his commission. Or maybe he was tired and thinking about the weekend ahead. Whatever his agenda or mood, I had to deal with it — and I was the customer.
I see a world where machines are actually more human than we are.
Now imagine this: I walk into the auto dealership and the salesman knows my name, my habits and even my mood. He can tell from the look on my face, the nature of my gait and the way I speak that I’m the type of customer who won’t haggle or compromise to get what I want. The color I’m looking for isn’t available? No problem, he says, here are your options — and he ticks off the cost of a paint job or vinyl wrap at several local body shops and helps me understand the opportunity costs of either solution. When I get frustrated, he knows it and shifts gears. He’s patient and compassionate. Now we’re talking about ordering the car in the color I want and waiting, months perhaps, for it to be delivered. I may not have exactly what I want right away, but that’s okay. I’m prepared to wait — and, more important, I’m happy with the outcome. It’s Customer Service 101. But here’s the hard truth: no company today provides the kind of service that I’m describing, as commonsensical at it sounds, consistently across its entire customer base. The technology isn’t there. Also, computers don’t get commissions — they can wait forever.
A better brain
There are scientists who believe that today’s technology — often called classical computing — will one day solve many of these customer service deficiencies. I think they are right, to a degree. The problem is, to provide a level of service that is consistently seamless, instant and malleable to whatever a customer thinks or feels at any given moment requires a level of computing power that current technology can never achieve.Here is where the promise of quantum computing comes in. Quantum computing is based on the principles of quantum physics, and the counterintuitive idea that any particle can exhibit two properties at once. For example, quantum physics holds that a single object can exist simultaneously in two separate places. In traditional computing, code is written in bits of information that are either 1s or 0s. Classical computers follow linear patterns of 1s or 0s, which means they can make only one calculation at a time. Quantum computing, however, allows a bit of information to operate as a 1, a 0, or something in between simultaneously. Thus, quantum bits can perform exponentially more calculations simultaneously to arrive at an answer faster than any modern computer. Time magazine recently did a beautiful job of putting the difference between conventional and quantum computing into layman’s terms. If you set out to find the lowest point in a mountainous landscape, they wrote, “a classical computer would do it like a solitary worker who slowly wandered over the whole landscape, checking the elevations at each point, one by one. A quantum computer could send multiple walkers at once swarming out across the mountains, who would then all report back at the same time. In its ability to pluck a single answer from a roiling sea of possibilities in one swift gesture, a quantum computer is not unlike a human brain.”
Let’s consider what this level of computing power could do. Artificial intelligence promises to change everything about the way we live and work. Couple that with the rise of quantum computing, and you have a pairing that is as portentous as the discovery of fire or electricity.
Our hybrid existence
How does a technology this powerful reinvent customer service as we know it? A human brain relies on a vast and complex network of neural nets, or classifiers, to work — say, to conclude whether another person is merely sad or disappointed or somewhere in between. To provide a degree of customer service that is almost intuitive requires an equally complex structure, instant and error-free. Classical computing can never achieve this. There are simply too many variables to consider in order for a computer to know that someone who is far away is the same person up close, or that someone whose head is turned and whose face is scrunched up is the same person looking straight ahead and smiling.
A customer calls on a Friday afternoon, tired. Imagine machine intelligence that will detect that, listen patiently and find a solution.
One of quantum computing’s most promising applications as it applies to sales and marketing has to do with inspiration. What galvanizes one person to buy a new smartphone while another passes it by? For companies, understanding the unique factors that motivate each individual customer to engage when she’s happy or sad or tired or curious is pure gold. Our human brains have detectors that signal to us what another person’s tone of voice or body language conveys. Conventional computers can’t do that. But quantum computers will, by virtue of their ability to process hundreds or thousands of variables at once. A customer calls on a Friday afternoon, tired. Imagine machine intelligence that can instantly detect that, listen patiently and help find a solution that works. Instead of two egos, two sets of preconceived notions or two different moods coming into play, there’s just one human dealing with a machine that’s 100 percent reasonable and fair. That’s the level of interaction — better than human — I want when I go to the car dealership or to the doctor.
The future of quantum computing doesn’t have the makings of a Hollywood blockbuster, thankfully. I don’t consider large-scale machine intelligence, even at the inevitable point when robots are sentient, to be a threat to humans. I see artificial intelligence as the inevitable next step in our evolution. We’ve been on this planet for hundreds of thousands of years. What makes us think we’re at the endpoint and not on the path toward some hybrid existence where machine-like features are embedded within us — and help us overcome challenges that are both primal and quotidian? I agree with Robert Jastrow, one of the 20th century’s great physicists and futurists, who predicted the next stage in human evolution will be “a new era of silicon-based life.” AI applications will show us the way.