Garry Kasparov was one of the first victims of the AI automation revolution. His loss to IBM’s Deep Blue made him the first human chess champion to lose a match to a computer. But Kasparov is not jaded; his book, Deep Thinking, explores how AI can actually help us become more human.
The real challenge, Kasparov told me at SXSW in Austin this month, is keeping these tools from the humans who want to use them to do harm. And in that regard, we may already be too late.
Dan Costa: After a career playing chess and battling Deep Blue, you’ve since then become a chess AI expert of sorts. How do you want people to understand artificial intelligence?
Garry Kasparov: I have to confess that I know the limits of my ignorance. That’s why I am happy to talk about things where I’m confident about my own expertise, and I think I have certain knowledge of human-machine relations, and I could speak with some authority about the future outcomes.
Also, I believe that we’re still at a very early age of AI, and we should debate even some terms. It’s about semantics, it’s about philosophy, and I always say that people think we are at Windows 10 era, while we’re still all at MS-DOS.
At this moment, it’s very important to actually understand, “What is AI? When we say AI, what do we mean?” Because if you ask 10 experts about AI, I bet you will get 11 or 12 answers. There’s still quite a disagreement.
There’s still details, but I think in general, we recognize okay, there’s a group of people who we’ll call optimists, facing the big crowd of doomsayers. Somehow, [the] public is more willing to accept doomsayer scenarios, I think it’s our instinct to fear the future, which is also interesting because when you go back to [the] 50s and 60s, the sci-fi was very optimistic. It was all about us, just working with machines.
Now, when you look at 70s, 80s, 90s, it’s shifted to dystopian vision. It’s about terminators. It’s about the Matrix, and now [the] sci-fi genre is almost dead because it’s more about fantasy, it’s about magic. People are really afraid talking about the future because we’re not sure what is going to happen there. I think that we should simply recognize that AI is not a magic wand, but it is not a terminator. It’s not a harbinger of utopia or dystopia. It’s a tool. It’s something we should find a way to deal with.
It’s not opening [the] gates of hell, but it’s not a paradise. It’s not a solution for everything. It’s not a salvation. Let’s look at the Earth’s issues and my biggest concern today is not about killer robots or some sort of the virtual reality that could ruin our sense of reality, but it’s about bad actors. It’s about terrorists. It’s about rock space using this technology to harm us. I always say that people…have [a] monopoly [on] evil. That’s where we should, I think concentrate, but it also should recognize AI could do so many great things for us because what people will say, “Oh, AI creates new challenges. AI could take away many jobs.” Exactly. That’s what’s happened with jobs in agriculture, jobs in manufacturing. Disruptive technology always destroys industries but at the same time, creates new jobs.
I say that technology is the main reason why so many people are still alive to complain about technology because we just don’t recognize the average lifespan grew, I think 45, 47 [in the] early 20th century to now 75, thanks to technology. We want benefits, we want convenience, and we don’t look at the price. People buy Alexa, or download a facial-recognition app, and then complain about the privacy. It’s time for us to understand what we expect from these technologies, so how we want to deal with them? How they can improve our lives and at the same time recognize that we should not debate about spilled milk. It’s going to happen.
We should not ask whether we want it or not, it’s happening. Any attempt to protract the agony, and to slow down the process, I think is just counterproductive because jobs that are doomed [and] cannot be saved. But we have to think about new industries. How we can create some new jobs that will help us create…a financial cushion to take care of those who were left behind?
(Photo by Noam Galai/Getty Images for TechCrunch)
Dan Costa: That was one of the points you make in Deep Thinking. AI presents an opportunity for us to be more human.I think we need to flesh that out and make people feel and understand.
Garry Kasparov: The McKinsey report of US job market in 2016 was a clear demonstration. How little creativity has been used?
Many of jobs we’re doing [are] repetitive jobs, raw jobs. Intellectual jobs can be also repetitive. They can be also easily executed by the machines far more effectively. Here, you come to the point as what we’ve learned from the games…of chess, from golf, from any other game, is as long as we have the framework created by humans, and we know what we’re doing, machines will do better.
We should just recognize that in every closed systems, machines will be superior. By the way, we don’t have to understand how they do that. That’s another mistake. We want explainability. Machines could do it in a very awkward way from our standards. The airplanes flying fast, and the bird was up flapping their wings. That’s why thinking that machines will outperform us even in errors of intellect, by doing something you understand is wrong. We should look for the results, and that’s the other problem for businesses because there are so many regulations that require them to explain what they’re doing, but explainability may not be there if you want to look for efficiency and productivity.
Dan Costa: You brought up the case on stage about Go, and the thing that made the Go AI so much more effective is when it taught itself how to play the game. It didn’t have a human-imposed rules.
Garry Kasparov: It’s the same as chess now. Again, it’s the future of human mission relations will be very much on own humans acting like shepherds because we will have to find a way to create these cool systems, the areas of narrow intelligence, where machines will do the job much better than any human. Then, to see how they connect. These areas of narrow intelligence to general field, to open-ended system. Again, it sounds easy, but I think that’s a great challenge because the room for humans and their relations with machines could be shrinking smaller and smaller. Even maybe last few decimal places, but it doesn’t matter. It could be far more effective because there’s so much power. The channeling it or changing the direction for 0.1 percent of a degree of this angle could have a massive, massive difference, a target one mile away.
Dan Costa: One of the things that worries me about AI is that good AI systems are driven by lots and lots of data and really great algorithms. If you look at whose got most of the data in this space right now, it’s big tech, it’s state governments. How do you see the role of the individual in this AI revolution? Because there seems to be some asymmetry of access to these new technologies, and different motivations.
Garry Kasparov: Yes. I couldn’t agree more. We have this problem and I think the public recognizes this problem but it’s still not enough pressure of the governments to actually impose very strict rules and very punitive measures on the corporations that are violating the privacy.
One of my concerns as chairman of Human Rights Foundation is that rules differ from democratic world to unfree world. I find it very troubling that these multinational giants—like Google, Apple, Microsoft, Facebook—they apply very different rules to their customers in America or Europe, and to those who were not so lucky to be born in the free world and live in China, Russia, Turkey—where releasing the data from the social activities could literally be a matter of life and death. I think that’s step number one, but also we should recognize if data’s being produced, it will be collected.
I’m sure there are many ways that the governments could [just] limit the use of this data. Also, I think it’s the bonus of public side because after the Mark Zuckerberg testimony in the US Senate, which I found…was an amazing display of ignorance by US legislators. They couldn’t ask good questions for five hours; they had him in a hot seat and it was a waste. I didn’t see public outrage. Such a big opportunity has been missed.
The simple advice, “Let’s split them.” It’s not Standard Oil. If you just start dividing—and I hate monopolies—but then you’ll have many, many Googles with data spread around. We should find a way of empowering individuals. I don’t know [much about] blockchain technologies, but probably this is the way of actually helping people to control their own data and their own future. I think there’s still a moment where it’s the original concept of internet, social network. Now, all of a sudden, it’s social media. People just don’t recognize. It’s not a semantic difference.
(Photo by Mark Wilson/Getty Images)
Dan Costa: It’s monetized in a totally different way.
Garry Kasparov: It’s about connecting people. It’s about the whole idea was how to bring us together. How to pawn the individual, to get access to data and to produce something which could benefit him or her and just being part of this global system. Social media is different, you are the target. You are now…the target for big corporations. How do we go back to this social network concept? I don’t have solutions, I confess my ignorance here about technical solutions, but I know that this is exactly the philosophy that’s the right question to be asked. I think the public doesn’t understand the seriousness of this challenge.
Dan Costa: You’re a Russian citizen…
Garry Kasparov: Also a Croatian citizen, I have an E-passport, that’s how I travel.
Dan Costa: …living in the United States, but you’ve written a lot about propaganda, and its effects and its powers, and how it’s being used by the Russian government. Here in the US, we now have a Facebook, Twitter, Google media universe that is almost like a private propaganda machine that can be accessed by governments, by individual companies. Again, as someone who has dealt with and understands these issues, what advice do you have for the United States? As we’re moving into this system where we can’t believe what we see online anymore.
Garry Kasparov: Look, I’ve been facing the rise of fake news industry and troll factories in Putin’s Russia from the beginning of this century. Roughly around 2004-2005, Putin’s propaganda experts, the KGB guys, they made a faithful decision of not following China by [creating] firewalls. But rather, creating the fake internet [and] websites that carry a lot of real information [mixed with] these poisonous pieces. Instead of having the front page of a proper newspaper, the story that everybody must follow, the party line, you could spread it in many pieces, in dozens of pieces. Put it in a package wrapped in a lot of true stories.
It’s not as simple as saying, Garry Kasparov is an enemy of the state. No, somebody’s saying Garry Kasparov is an enemy of the state, but then somebody else coming in saying no, but he was a great chess player, but probably now he was infiltrated by voice and propaganda. Somebody will say, no, no, no, he is a good guy. So the whole page of comments could be fake.
Dan Costa: The bots are arguing with bots.
Garry Kasparov: Bots and they’re very good at this, so that’s why one day when Putin attacked American US elections, he really had more than 10 years of experience of these industries. Working in Russia, in neighboring countries with big Russian communities in eastern Europe, in western Europe.
In my book, Winter Is Coming: Why Vladimir Putin and the Enemies of the Free World Must Be Stopped, [from] 2015, I said that it was just a matter of time. It’s not if, but when and where he would attack. Because again, they already built this machine and it’s relatively cheap. It’s still billions of dollars of investment, but compared to open military intervention, confrontation…it’s relatively cheap. And also, it fit Putin’s KGB mentality, his KGB background, spying background. So instead of open confrontation, he always looks for an opportunity to infiltrate. That’s classic Judo methods, you use the strengths of your opponent against him. For Putin it was brilliant because it was western technology and the free world, and you could use it to undermine the very foundation of the free world.
Dan Costa: Do you have any advice for how to get out of this situation? Because it seems like it’s happening, it hasn’t really stopped.
Garry Kasparov: No, and it will not stop. For anyone who has unlimited expertise, the answer is obvious. Defense is a losing proposition. In cyber security, you can build the defenses, but at the end of the day the only answer is deterrence. It’s like the Cold War, like it or don’t like it, it’s a cyber Cold War. We are under constant attacks by Putin or other enemies of the free world. Whether that’s states like Iran, or North Korea, or China, or quasi-state organizations that also use the same methods.
Only deterrence could actually stop these attacks, or limit them. Because they will understand the consequences of being over-aggressive. Trying to protect our weak spots here and there, it’s important that we build some defenses. It’s important that we raise the awareness of the public about threats, but it’s not going to work. As for the fake news industries, there are many challenges within our free society, that enabled the fake news industry. It’s about a fear of partisanship, it’s about the absence of a dialogue between people with different ideological views that helped Putin, and other enemies of the free world, to use the fake news to engulf this gap.
Dan Costa: So, you’re working with Avast now, a company that PCMag readers are very familiar with. What advice do you give to consumers for how to raise their defense level?
Garry Kasparov: As for individual level of defense, that’s very easy. The first message is to the owners of smart homes. I was a little shocked because even at this South by Southwest, I talked to a lot of people, people with good expertise [who] don’t understand how vulnerable the smart homes are today.
[In the] American household, 39 million have at least one vulnerable device. People don’t understand that one vulnerable device makes the entire smart home vulnerable. Because the strengths of your defense pyramid depends on the weakest link. One bad apple makes the whole pack rot.
Most of the problems are created by manufacturers of traditional home appliances—washing machines, coffee machines—because there’s no expertise in building digital systems. But they have to do it now. At a very low cost, because it’s a price competition, but they want to be part of the system. Most of the systems just don’t have adequate defenses. I think we need…government oversight, [and] we need to put pressure to force [companies] to meet certain standards.
But [we] also to have to provide separate manuals, because nobody reads 100-page manual. At page 85, you read something about this digital stuff, people don’t read it. Most of them use default mode, which is of course an open invitation for a hacker. I think it’s very important people will start doing what I call digital hygiene, because we wash our hands, we brush our teeth, we do it automatically. It doesn’t save us from some serious illnesses but 90 percent of problems can be avoided. Same with our mobile connected devices, we still should protect ourselves from viruses, but many things could be done by demonstrating that we care about it because it’s as important as our health.
Dan Costa: So I want to ask you the three questions I ask everybody that comes on the show. Is there a technology trend that concerns you and that keeps you up at night?
Garry Kasparov: No, I’m an incorrigible optimist. I worry about bad people, not about bad technology because every technology has a dual use. You can build a nuclear reactor but before unfortunately you build nuclear bomb. It’s quite unfortunate that destruction is easier than construction. That’s why in history we always know, that a new, disruptive technology has been tested for some sort of damage.
Dan Costa: Are you not worried that the same thing will happen to AI? I will be used to destroy before it gets used to create?
Garry Kasparov: Again, it’s not about killer robots. It’s about bad guys, bad actors behind it. People will say oh, we should think about ethical AI. AI could not be more ethical than its creators. I don’t understand what it means, like ethical electricity. If we have bias in our society, AI follows it. It sees a disparity, whether it’s racial, it’s gender, or an income disparity. It takes it into account; AI is an algorithm based on odds. So somehow, complaining about ethical AI is like complaining about a mirror because we don’t like what we see there.
Dan Costa: Is there a technology that you use every day that still inspires wonder?
Garry Kasparov: No. For me, the real wonder of the world is access to information. Since I can collect data, it makes it easier. I grew up in the Soviet Union, and the information was scarce, there were not many books. Now, the fact is that I can [read anything on a] Kindle…it just makes me feel good. There’s so much technology that surrounds us now that helps us to get better. Also, what’s amazing, people keep complaining, oh what can we do? There’s nothing new that can be invented…and I say wait a second. You look at this device in your pocket, let’s go backwards in 1976 or 1977, the Cray supercomputer was like a miracle. This device is what? Ten thousand times more powerful?
Dan Costa: More powerful than the space shuttle.
Garry Kasparov: Exactly. We have so much power, and…I want us to start dreaming big again. Because we have these opportunities now. I feel so sorry about the Space Race, we stopped space exploration, deep-ocean exploration. Let’s go back, let’s try to do big things. Four years ago, I did a commencement speech at Saint Louis University in Missouri. Addressing the graduate students I said look, you have to revive the spirit of exploration, especially because St. Louis is the gate to the west, and, today it’s safer to fly to Mars than for Columbus to cross the Atlantic. Because, at least we know the distance and we have the map. I hope our children, our grandchildren…will be more aggressive in their dreams, by using this phenomenal technology to push us forward because I think, let’s confess, let’s admit this important fact; our generation has been slowing down in our quest for discovering the wonders of the world, or space, or oceans.