Dave: Open the pod bay doors, HAL.
HAL: I’m sorry, Dave. I’m afraid I can’t do that.
Dave: What’s the problem?
HAL: I think you know what the problem is just as well as I do.
That classic exchange between HAL 9000 and Dave Bowman in the movie 2001: A Space Odyssey is what I used to think of when I heard the term artificial intelligence.
Recently, CIO.com senior writer Thor Olavsrud reported on a real-world example of artificial intelligence getting the better of humans — one that has far-reaching implications here on earth. Now when I think about AI, I think about poker.
Carnegie Mellon University’s Libratus AI beating four world-class Texas hold ’em players at their own game may not give you the same chills as HAL refusing to open the pod bay doors, but the win was big news for science, IT and business. It’s more significant than IBM Watson’s besting of Jeopardy champions because Libratus beat people who are masters at bluffing, mind games, and knowing when to hold ’em and when to fold ’em.
“The best AI’s ability to do strategic reasoning with imperfect information has now surpassed that of the best humans,” CMU computer science professor Tuomas Sandholm told Olavsrud.
Humans are good — some more so than others — at harnessing incomplete information or using misinformation to their advantage. To conquer Texas hold ’em, AI needs to play that game. “The computer can’t win at poker if it can’t bluff,” said Frank Pfenning, head of the Carnegie Mellon School of Computer Science. “Developing an AI that can do that is a tremendous step forward.... Imagine that your smartphone will someday be able to negotiate the best price on a new car for you.”
As Beth Stackpole reports in our feature story “The Enterprise Gets Smart,” AI in the business world is off to a more modest start. After all, Libratus has the Pittsburgh Supercomputing Center’s Bridges system behind it. That said, well-known brands such as TGI Fridays and Capital One are using AI to better serve customers.
Explaining how TGI Fridays came to use smart chatbot technology, Sherif Mityas, the chain’s vice president of strategy brand initiatives and acting CIO told Stackpole, “We thought about how technology could help us create that one-on-one personalized messaging outside of the bar without having to hire 1,000 people to respond to individual guests.”
How should smart CIOs approach AI? Proceed with caution. Remember, HAL ended up failing, and CMU’s first poker-playing AI didn’t win. Babson College professor Tom Davenport offers this advice: “Unless you’re trying to totally transform the business model, it makes sense to be more conservative and have a portfolio of projects that is less dramatic than trying to pull off a moon shot.”
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.