The Conversation: Is Artificial Intelligence a Virtuous Reality?

Featured image:
© Isaac Lawrence/AFP/Getty Images

Through advanced technology, humans are getting smarter and living longer, and machine learning, artificial intelligence and virtual reality are all becoming part of the world's lexicon. Ray Kurzweil, a futurist and director of engineering at Google who heads up a team developing machine intelligence and natural language understanding, sat down with Andrew Ross Sorkin, a New York Times columnist, at the Global Leaders' Collective in Washington, D.C., convened by The Times in November 2016. They discussed A.I. and its critics, population sustainability and how we may soon be able to connect our brains to the cloud. An edited and abridged excerpt from their conversation follows.

Andrew Ross Sorkin: You talk about the idea of physical immortality and that this is going to be possible by 2045. Can you explain?

Ray Kurzweil: I'll never be able to say I've lived forever because it's never forever, but I'll talk about three bridges to radical life extension.

Bridge one is what you can do right now to stay healthy, the old-fashioned way, so that we can get to bridge two. A key idea is that information technology progresses exponentially, and health and medicine are now information technology. For example, you can now fix a broken heart — not yet from romance, that'll take a few more developments in virtual reality. We can now fix the damaged hearts of heart attack survivors by reprogramming adult stem cells. We're regrowing organs, successfully installing them with the patient's own DNA in animals, and we'll be able to do that in humans soon. What's now a trickle will likely be a flood within 10 years. That's bridge two.

That will get us to bridge three, where we'll have medical nanorobots — little robots that are computerized, the size of blood cells — basically finish the job of the immune system. We have blood cells, our T cells, that keep us healthy, but they evolved [back when] there wasn't any interest in the human species for us to live very long. They don't recognize cancer, for example, because that gets us later on in life. We can finish the job of the T cells with these medical nanorobots. There are detailed designs of how to go after every disease once we have these devices. That's a 2030 scenario.

Ultimately we're going to merge with artificial intelligence.

Andrew Ross Sorkin: The "singularity." Part of the future, in your mind at least, is something known as the singularity.

Ray Kurzweil: We start with the idea of extending our mental capacity with artificial intelligence. Most people, if they leave their cellphone, they feel like they're incomplete. They're not yet inside our bodies and brains — although there are some people that have computers in their brains, like Parkinson's patients — but that'll be routine in the 2030s. Another application of these medical nanorobots will be to connect our neocortex — that's the outer layer of the brain, that's where we do our thinking — to the cloud.

We'll connect our neocortex wirelessly to the cloud just the way your cellphone does it. We will be a hybrid of biological thinking and nonbiological thinking, which I believe has already started with these devices outside our body, and we'll become smarter. By 2045 we'll expand our intelligence a billionfold. It's such a profound transformation that we borrow this metaphor from physics and call it a singular change in human history.

Andrew Ross Sorkin: Let me give you two other smart fellows who have a less rosy version, this relating to A.I. Your friends Elon Musk and Bill Gates say that A.I., artificial intelligence, represents our "greatest existential threat."

Ray Kurzweil: I wrote in my 1999 book, "The Age of Spiritual Machines," about the grave intertwined promise versus peril of artificial intelligence, nanotechnology and biotechnology. I talk about three phases that people go through when they look seriously at the potential of these new technologies. One is inspiration that these technologies have the potential to overcome age-old problems like poverty and disease and short life spans. Then fear at the potential dangers of these technologies. And finally a balanced view that these existential dangers exist, but we have means of overcoming them.

Audience question: As these other parts of the world evolve, how can our emotional intelligence evolve as well?

Ray Kurzweil: We still have that old brain that provides basic motivations. The neocortex, which is around the old brain, is really the great sublimator. I may have ancient motivations for aggression and conquest. My neocortex will sublimate those into writing a book about the future or talking at a conference to leading fashion executives. No other species does these types of things. The neocortex is organized hierarchically. At the very bottom I can tell that that's a straight line, at the top I can tell that something is funny or ironic, or that someone is pretty.

We're going to add additional levels to the hierarchy when we can actually expand our neocortex by connecting to [a] synthetic neocortex in the cloud. We will become funnier; we'll be better at expressing loving sentiments. Those types of emotions, the ones that we regard as the finer qualities of humans, exist at the top of the neocortical hierarchy. And we're going to enhance those as we increase our brain capacity.

Audience question: The idea of radical life extension is exciting, but how are we going to solve issues like resource utilization and sustainability with an increased population?

Ray Kurzweil: We have far more resources than we need. We have exponential growth, for example, in solar energy. Once we meet all of our energy needs from solar, we'll be using one part in 10,000 of the sunlight. It's a similar story with geothermal, tidal energy, wind and so on. We have thousands of times more energy than we need. Vertical agriculture will provide food at very low cost for the entire population. We'll ultimately be able to print out the other physical things we need with advanced A.I.-controlled manufacturing technologies in the 2020s to meet the material needs of the population.

© 2017 The New York Times
Distributed by The New York Times Syndicate