What it is: AlphaGo is a machine learning program that has mastered the complexity of the ancient game of Go.
For the longest time, people believed that it was impossible for a computer to beat a grandmaster at chess. Then IBM’s computer, called Deep Blue, beat Garry Kasparov after losing the first match. After computers have routinely beaten human chess masters, critics then claimed it was impossible for an artificially intelligent program to beat anyone at the game of Go.
Unlike chess, Go has far more possible legal moves to consider, making brute-force calculations impossible. Yet Google decided to create a Go-playing AI program called AlphaGo with the sole purpose of defeating a Go champion. In 2014, AlphaGo took on Fan Hui, a highly regarded European Go champion, and beat him 5-0. Now AlphaGo has its sights set on challenging Lee Sedol, a top ranked Korean Go champion, to a match.
Whether or not AlphaGo succeeds in this match is irrelevant. The fact is that each time there’s been a challenge that people believe AI could never solve, AI has gone out and solved that problem. When that happens, critics simply move the goal posts and claim AI can’t solve a new problem — right up until the day it actually does.
In the early days of automobiles, people rightly claimed that horses were more reliable and faster, and they were right — right up until the time they weren’t. When you believe something is impossible, it’s a safe bet that you’ll be proven wrong eventually. For advanced technology like AI, impossible problems of today will likely be solved tomorrow far sooner than you might believe possible.
As long as computer scientists believed it was possible to create a Go-playing AI program that could learn from its mistakes and get better, the seemingly impossible goal of creating a Go-playing program became possible. That’s just the way technology advances so when you think today’s technology has no reason to change, you’ll be blind-sided when it does change. Just witness the massive shift from PCs to smartphones and tablets.
AI is getting smarter and critics of technology remain at the same level of ignorance and stupidity that they’ve always been stuck at. When you look at the progress of AI and technology, and the lack of progress in the minds of so many critics, it’s easy to see that machines really can learn while so many human beings simply cannot.