The buzz around artificial intelligence is hard to miss. For the most part, we now expect that artificial intelligence will be a part of our lives in some shape or form. Does that mean that we want to be riding around in self-driving cars five years from now? Not exactly. The reason is that, while we accept AI as part of our lives, we still consider it an imperfect technology prone to error. As with anything else in life, there is a good and bad.
The good. Over two weeks ago, we heard about Google’s AI out-maneuvering Lee Sedol, a grandmaster of the 2,500-year-old, complex game called Go. Sedol lost three games straight to Google’s AI before defeating it in the fourth game. Ultimately, Google’s AI won a fourth and final victory in game five despite making a huge error at the beginning of the game. The potential good in this example of artificial intelligence is that besides navigating a complex game, Google’s AI was able to dig itself out of a mistake that ultimately led to success against a worthy opponent.
The not so good. Just last week, Microsoft’s AI chatbox known as Tay was launched on Twitter and went from reasonable tweets to straight chaos in less than 24 hours. The smart-learning software is intended to engage people in playful and casual conversation. However, what the AI learned from interacting indiscriminately on Twitter wasn’t necessarily great. Those who shared misogynistic or racist comments with Tay were later repeated back in mass creating the need for Microsoft to go through and manually delete some of Tay’s comments. This is just one example of how AI can backfire. As they say, if you put garbage in; you get garbage out.
Again, artificial intelligence is here to stay, but just like the robots and programs themselves, none of us should follow them blindly. There is something to be said for the human characteristics of empathy, common sense and humor.