When Garry Kasparov was defeated in a game of chess by IBM’s Deep Blue in 1997, it represented more than just a PR exercise from an opportunistic computer manufacturer looking to bump its share price.
For many, it signified the start of the end of human supremacy: the machines had won.
Today, 20 years later, in certain aspects, the human mind’s inferiority to machines is an accepted fact.
Now, countless operations better suited to neural networks and automated processes exist, from cancer detection to mundane back office tasks. Like all cycles, a new sequence has begun. We’re back in much the same position we were in 1997. Fears of artificial intelligence (AI) and automation are front page news.
Our appetite for the unknown is matched only by our thirst for fear.
Kasparov is in London with Avast, an information security firm for which he is ambassador, speaking at IP Expo.
Forehead swollen and bandaged, the chess world champion of 15 years is three hours late for our interview due to an earlier car accident, but in good spirits, considering.
Since retiring from international chess in 2005, his life has been a perfect storm of human rights activism and anti-Putin polemic. Alongside his political life, his unique relationship with machines – after two decades of reflection – has turned Kasparov into a poster boy for embracing the future. Who better to discuss AI with, than a man who was famously beaten by it?
“I recommend being very careful using the word AI,” he insists, in a gravelly Russian baritone. “If you have 10 experts in the room and ask how they define AI, you may end up with 15 different answers. I think we’re still at the stage where we’re yet to define what AI is, and you have to start with the ‘I’ – intelligence.”
Kasparov says it’s a common misconception to put Deep Blue in the bracket of AI. Brute-force computing capable of 200m positions per second is a maddening and worthy opponent that demonstrated the game of chess could be cracked. But the software follows a rational path. It does not learn – so is not intelligent. A modern equivalent, such as Google’s DeepMind, makes its own decisions in ways humans can’t comprehend.
Middle class machines
For a man whose public battles with a machine were seen as a harbinger for a new era of computational prowess, he remains sanguine, insisting machines are not our enemies.
“It’s just a new step in the history of human progress. From the very beginning, we came out with new concepts, new machines, new tools to help us to do repetitive tasks and concentrate on something more productive. And for hundreds, if not thousands, of years, this process was never questioned. But the difference today,” he chuckles, “is that for the first time machines are threatening the jobs of people with college degrees and Twitter accounts.”
Given the perceived inevitability of job losses due to automation, some (neo-luddites if you will) are questioning whether everything that can be automated should be automated. Fears of creating a political “underclass” – leaving those without a digital paddle drowning in the tide of automation – are not unfounded. Research out of the world’s top universities, from Oxford to MIT, offers varying volumes of exorbitant cross-sector job losses. Kasparov says that, rather than discussing whether processes should or could be automated, humanity must accept that they will be.
I think it’s totally counterproductive to discuss whether things that can be done, will be done.
“I think it’s totally counterproductive to discuss whether things that can be done, will be done. They will be! What does it mean for us? Now that’s a different story. Go back to chess. We have plenty of experience showing how machines and humans can work together to improve dramatically the effect of decision making.
“And that’s why I remain optimistic. I don’t think machines will ever reach 100 per cent perfection – which means that there’s always room for humans to compensate for the remaining deficit.”
Kasparov goes on to argue that any attempt to slow down this cycle of innovation will not destroy the cycle, but keep destroying jobs without creating new ones. “The whole idea is that you generate enough new income through these industries to find a way to help those who have been left behind,” he says.
State of affairs
Governments around the world have been sluggish to move on technology – whether through lack of understanding, or through forlorn attempts to preserve the status quo. Governments and citizens alike are ill-prepared for the approaching revolution. Usually when the state doesn’t understand something, it gets swept under the regulatory rug – such as our current government’s attempt to ban end-to-end encryption.
“As someone who was born and raised in the Soviet Union,” says Kasparov, “I’m very suspicious of government authority. Especially when government tries to interfere with science and progress. If something is doable, it will be done. And if you try to stop it here, you have the Putins of the world who will buy it and use it for very destructive processes.”
Some analysts predict that due to the lack of freedom in a state such as China, the data the government can collect is much richer, so its technology will advance more quickly.
I hit a nerve when I ask Kasparov if he believes there is an AI arms race.
“I’m surprised people still ask me this question,” he says, pointedly. “The free world always came up with the most sophisticated and more advanced technologies, simply because in non-democratic systems, you have so much bureaucracy that stands in the way.
“Mathematically it’s more difficult. You can have 100 different Silicon Valley startups, but there will be only one Steve Jobs. But you don’t know who the Steve Jobs will be. Centralised systems cannot programme failure. If you look for something to break through like Apple in 1977, you have to start programming 99 failures.”
Artificial intelligence and automation is an evolutionary step, as humans augment the physical with the digital realm. Kasparov is optimistic for a future in which humans and machines work side by side, but is concerned by public perception of the unknown.
There was a time, we agree, that SciFi movies were all about exploration, travelling space or Atlantis, discovering new worlds and curing the ills of ours.
At some point, a switch flicked in America’s mind.
Dystopia was the new black: the Matrix made us question reality, apocalypse was now. Today we have Elon Musk and Stephen Hawking, people who should know better, injecting the public sphere with a fresh dose of fear of the future.
“It’s a self-fulfilling prophecy,” Kasparov sighs, slapping his hands on the table. “If you think it’s dark and bleak, you will end up in darkness. But I think it’s also the demand. And demand has been growing over decades of people getting more nervous about the future.
“I think it’s quite ironic that Elon Musk, a great innovator, is coming up with these stupid predictions. We don’t know what the future is!
“If you keep repeating negatives, you lose sight of the positives, because every technology is neither good nor bad – it’s agnostic. You can use it in many forms. It’s the same with nuclear energy. You can make a nuclear bomb, or you can make a nuclear reactor.”
Elliott Haworth is business features writer at City A.M.