“It isn’t looking good for humanity.”
That’s how the New York Times article on AlphaGo defeating Chinese Go Grandmaster Ke Jie began. A year after AlphaGo’s convincing victory of Lee Sedol in Seoul, DeepMind’s artificial intelligence software also brought the world’s number one player to his knees.
Considering Mr. Ke’s public remarks before that AlphaGo can’t beat him, this victory adds more fuel to the public fear that artificial intelligence will begin to replace human labor in more aspects of life.
But putting dystopian fears and universal basic income discussions aside, AlphaGo’s victory provided a glimpse at the future of creativity, a future driven by machine intelligence.
Extracting and Discovering New Knowledge with Artificial Intelligence
Perhaps blinded by the effects of automation as a result of the growing adoption of machine intelligence, we often forget the promise and potential of artificial intelligence: the ability to extract and discover new knowledge. One understated achievement of AlphaGo is proving that machines can, in fact, be creative. See the following excerpts from NYT and DeepMind’s coverage of the event [emphasis mine]:
“AlphaGo…has already pushed assumptions about just how creative a computer program can be. Since last year, when it defeated a highly ranked South Korean player at Go, it changed the way the top masters played the game. Players have praised the technology’s ability to make unorthodox moves and challenge assumptions core to a game that draws on thousands of years of tradition.” — NYT
The emphasis on creativity is echoed by DeepMind’s thoughts on this milestone:
“The creative moves it played against the legendary Lee Sedol in Seoul in 2016 brought completely new knowledge to the Go world, while the unofficial online games it played under the moniker Magister (Master) earlier this year have influenced many of Go’s leading professionals – including the genius Ke Jie himself. Events like this week’s Pair Go, in which two of the world’s top players partnered with AlphaGo, showed the great potential for people to use AI systems to generate new insights in complex fields.” — DeepMind
Artificial Intelligence: Beyond Just Games
While the algorithms behind artificial intelligence systems remain fairly domain specific for now, AlphaGo has proven that its algorithms can not only challenge status quo in human thinking of the topic, but also teach new and creative approaches to solve complex problems.
DeepMind has released games that AlphaGo has played against itself and other top players to provide insight into how a machine analyzes and thinks about the game. We are already reaping the benefits of this as Mr. Ke, in the match against AlphaGo, seemed to have made moves inspired by AlphaGo’s playing style.
Admittedly Go, at the end of the day, is merely a game. But imagine if we can generalize AlphaGo’s success in other domains to tackle our biggest problems, such as “finding new cures for diseases, dramatically reducing energy consumption, or inventing revolutionary new materials.”
Creativity and Artificial Intelligence
In one of the most cited papers on creativity and artificial intelligence, Margaret A. Boden claimed that “AI techniques can be used to create new ideas in three ways: by producing novel combinations of familiar ideas; by exploring the potential of conceptual spaces; and by making transformations that enable the generation of previously impossible ideas.”
Perhaps in the near future, we can even combine all of these approaches to model unique ways to solve a problem, borrowing from other domains for innovation. More broadly, maybe creativity has always been like this.
For ages, humans have borrowed from nature to inspire new products such as teflon and velcro. With artificial intelligence, this process can simply be accelerated with computers’ ability to process millions of possibilities.
This potential of artificial intelligence brings me to an interesting interview with Garry Kasparov, the great chess player who became famous for his match against IBM’s Deep Blue back in 1997.
Back when Deep Blue beat Mr. Kasparov in chess, the world had a similar reaction to what we’re seeing now with AlphaGo. But to Mr. Kasparov, the important takeaway isn’t that machines are taking our jobs away, but that the combination of humans and AI can be more powerful.
Yes, AI is playing an uncomfortable role of “breaking up the ice of complacency” and “threatening too many comfortable jobs” as Mr. Kasparov states. But machines cover human weakness that center around psychological aspects of solving a problem.
In chess, for example, a human would rarely give up a queen, even if in two moves sacrificing the queen now will win the game. The machine, on the other hand, will always optimize for the best result and ignore the psychological impact of such a move.
In some domains, like chess, the computer’s approach may be superior. But in other domains, such as healthcare, the human intuition can be life-saving or morally correct over the machine’s calculation.
Mr. Kasparov argues that machines and humans working together brings about a stronger combination. With artificial intelligence now unlocking logic-based creativity, the question becomes will machines force humans to raise their level of emotional intelligence to counterbalance the emotionless decisions of computers?