SAN FRANCISCO (CN) – Garry Kasparov, the former World Chess Champion dethroned in 1997 by an intelligent chess-playing computer, said Thursday that people must discard their “mythological” fear that artificially intelligent killer robots will annihilate humanity, and collaborate with AI to advance scientific discovery.
“How about bringing together the best of two worlds: human creativity, our fantasy, our ability to strategize, and machine brute-force calculating,” Kasparov said in a keynote address at the Train AI conference in San Francisco. “It is time for us to start reconsidering these relations and find the best way of incorporating with machines.”
Artificial intelligence – computer systems capable of intelligent behavior – is becoming ubiquitous. Today, it is found in a range of applications, from search engines and spam filters to automatic speech recognition software and medical image analysis tools. The technology is being incorporated into self-driving cars and drones for disaster relief operations; further in the future, it could speed scientific research and help develop cures for deadly diseases.
But some experts fear that if artificial intelligence surpasses human intelligence, it will drive us to extinction.
Cosmologist Stephen Hawking warned in 2014 that although the advent of superintelligent machines would be “the biggest event in human history, it “might also be the last, unless we learn to avoid the risks.” For example, a superintelligent system given the task of curing cancer might conclude that the only way to do so is to kill every person on earth. Free of human control due to its superior intelligence, the system would set about its gruesome task.
Kasparov, who writes and lectures on the intersection of human and machine intelligence, dismissed those fears on Thursday as “science fiction.” He instead envisioned a future in which humans and machines collaborate to make advances that would otherwise be impossible.
In Kasparov’s scenario, a human “operator” would set the framework within which a machine must solve a problem. If the operator doesn’t properly define the framework, the machine’s results are useless. Humanity’s preeminence is thereby preserved.
Jobs and industries will evolve and adapt; as machines become smarter, human operators will require less training and expertise to oversee them, Kasparov added.
“Each problem will require a combination of humans and machines; it will require very specific human qualities that will help us unleash human creativity,” he said.
But University of California at Berkeley economist Laura Tyson and global consulting firm McKinsey & Company estimate that 400 to 800 million people in low and medium-skill occupations will be forced to change jobs by 2030 due to displacement by intelligent machines, causing large-scale societal disruption. Meanwhile, AI will increase wages in highly skilled occupations, such as computer science and engineering, increasing income inequality.
According to Tyson, half of the activities that workers are currently paid to do can already be automated with available technology.
Other nearer-term AI-related threats exist. They include terrorists hacking into drone systems or self-driving vehicles to deliver explosives or cause crashes, or into servers used to direct autonomous weapons to attack civilians, according to a February 2018 study by researchers at Cambridge and Oxford universities, among others. Fake news reports can be disseminated using fake video and audio of politicians making inflammatory comments, and people in swing districts can be targeted with personalized messages to influence voting behavior.
“It feels like a bit of an arms race,” Lisha Li, a principal at venture-capital firm Amplify Partners, said in a panel discussion Thursday.
Despite the potential threats, Robert Munro, chief technology officer with Figure Eight, the artificial intelligence firm that organized Thursday’s conference, was optimistic about AI.
Before earning his doctorate in natural language processing at Stanford University, Munro said in a keynote address he installed solar panels on Liberian medical clinics for the United Nations. While there, Munro’s team heard rumors that a new group of refugees had come over the border from Côte d’Ivoire, and was camped out in the next valley. Due to fighting, the U.N. couldn’t get to them for two days.
And despite the fact that both Munro’s team and the refugees had cellphones, they couldn’t contact one another to learn how many refugees there were and what supplies they needed.
Now, the U.N. and Figure Eight are building an artificial intelligence tool to estimate population densities. The tool will use satellite and drone images to predict how many people are in a given area, so that aid workers will know how much food or other supplies people need.
“Humans are working with computers to solve this really important problem for humanitarian crises, which is something I wasn’t able to solve at all 10 years ago,” Munro said.