Artificial Intuition
The idea from this weeks readings that really got me thinking was the idea of artificial intelligence trying to mimic human intuition. We know that one of the biggest potential benefits of A.I. is its ability to process and utilize the massive amounts of data that is being collected today. However, what about intuition, that force we can’t quite put our finger on but is something responsible for guiding our actions?
Websters defines intuition as: the power or faculty of attaining to direct knowledge or cognition without evident rational thought and inference. Basically doing something not because the data supports it but because it just feels like the right thing to do. In the NYTimes article Meet the People Who Train the
Robots (to Do Their Own Jobs), they mention an A.I. name Harrison that can book travel reservations. However, when it comes to calling the hotel to try to get room upgrades or recommend how to get the most out of a vacation, “That’s something A.I. can’t do.” In How Companies Are Already Using AI
by Satya Ramaswamy, he talks about how he sees A.I tackling the low hanging fruit when it comes to jobs, automating production cycles and machine to machine interactions, rather than really performing human-like tasks.
This got me thinking, is human intuition really something indescribable that machines will never be able to mimic? Or is it something that can be boiled down to numbers and data if the technology is good enough? How would one go about training an A.I. to give them a human sense of intuition? One potential idea, as described in this article by MIT News, is to train them using very high skill level humans. By encoding the strategies of high-performing human planners in a machine-readable form they were able to increase the effectiveness of the algorithms in question by 10 to 15 percent. But is this really giving them human intuition? It makes logical sense that an A.I. trained by a highly skilled human would perform better than one trained by a moderately skilled human, but will that really translate into something similar to human intuition? I don’t know the answer, but it is interesting to think about. I wonder what other strategies might work for trying to train intuition into our A.I’s?
