AI is No Laughing Matter

Alexa can do almost anything you ask of it. Play your favorite song? Turn on your TV? Give you a weather report? Show you pictures of cats? No problem. Alexa can also tell a good joke, but it can’t laugh, even if you tell Alexa an even better joke. No, Alexa isn’t the greatest comedic “straight man” since Dean Martin, Dan Rowan or Dick Smothers (really dating myself here). Unfortunately, Alexa can’t laugh at jokes because it has no sense of humor. It simply doesn’t get the joke any more than your pet poodle or potted plant does.

Likewise, an autonomous vehicle can automatically and suddenly brake to avoid hitting another vehicle stopped in the middle of the road, but it has no idea that if it didn’t slam on the brakes a wreck would have ensued.

Artificial intelligence (AI) has its limits, especially when it comes to very human traits. Humor, irony, empathy, fear, jealousy, etc. are not AI strong suits. In fact, they literally aren’t in the program, which is why you’re not going to carry on an interesting, extended conversation with Alexa any time soon, if ever. At the same time, AI’s lack of emotion or bias makes it very effective in doing what it does best.

In this March/April 2019 issue of Analytics magazine, Joseph Byrum, our resident AI guru/philosopher/whisperer and member of INFORMS, weighs in on AI’s strengths and weaknesses in his feature article, “The Futile Quest for Artificial General Intelligence,” in which he asks: “You have to wonder, even with all the power now at our disposal, why are Alexa, Siri, Google Assistant and Bixby still not quite right?”

Hey, nobody’s perfect, not even AI. 

Rest of the Story

Speaking of artificial intelligence (aren’t we always in Analytics mag?), longtime INFORMS member Doug Samuelson reports on a new study by a leading Washington, D.C., think tank that’s engaged in defense and national security issues. The article, “Nurturing Artificial Intelligence,” summarizes the case for creating an “AI ecosystem” and why it’s critical for successful AI implementation. Meanwhile, data scientist Jim Theologes makes the case for “Neural Networks and Efficient Modeling,” the second of a two-part series.

Artificial intelligence and neural networks (sometimes described as a steppingstone toward AI) are closely related, date back decades and are now enjoying renewed interest after early setbacks thanks to the explosive growth of data and computer power.

Other highlights include N.S. Krishnan’s feature on “Culture, Talent and Technology,” in which he outlines the pillars needed to build an analytics-driven supply chain organization; John Oechsle’s “SMB Guide to CRM in 2019” that tells you all you need to know about customer relationship management systems; and our diverse lineup of popular departments such as “Analyze This,” “Healthcare Analytics” and “Five-Minute Analyst.”