Trending: Amazon employee at Seattle-area warehouse tests positive for COVID-19
Paul Misener
Paul Misener, Amazon’s vice president for global innovation policy and communications, talks about Amazon’s “invention machine” at the Global Artificial Intelligence Conference. (GeekWire Photo / Alan Boyle)

Taking advantage of artificial intelligence and machine learning may be part of Paul Misener’s job as an Amazon executive, but he’s doing it for fun as well.

Misener, Amazon’s vice president for global innovation policy and communications, gave a personal endorsement for Amazon Web Services’ machine learning platform today at the Global Artificial Intelligence Conference in Seattle.

“Amazon SageMaker is a really cool service offered by Amazon Web Services,” he told the audience at the Washington State Convention Center. “This brings machine learning out to everyone, including me. I’ve done some fooling around with things on it, some hobby things.”

Like what? We had to ask.

“My son and I have a project where we’re assessing the occurrence of tornadoes in the Plains states in the United States, and relating them to weather measurements in advance,” he explained. “So the ML is looking at the data set and saying, is it possible that a slight drop in air pressure two days in advance of a tornado .. is there some causality there? Could it lead to better tornado predictions?”

Believe it or not, there’s some evidence to link air pressure drops with subsequent tornado activity — enough evidence to justify sifting through big data.

Misener compared AI’s power to the pattern-finding prowess of Nobel-winning mathematical John Nash, as portrayed in the movie “A Beautiful Mind.”

“Remember in the movie, where he looks at the data? That was fiction, right? That didn’t happen. But machines can find those patterns, so that’s an application I use in everyday life,” Misener told the conference crowd. “Look, if I can do it, everybody in this room can do it.”

The point is that artificial intelligence and machine learning are likely to make their way into kids’ homework and everyday routines, just as home computing has.

“Machine learning used to be something that was available only to those who had supercomputers and a lot of big, brave data scientists to run them,” Misener said. “Not anymore. Artificial intelligence is now available to anyone who wants to use it.”

Misener’s excellent machine-learning adventure is just one example illustrating how Amazon and its executives approach innovation inside and outside the workplace. In the course of today’s talk, the 19-year Amazon veteran recapped many of the guiding principles for the Seattle company’s “invention machine”:

  • Obsess over the customer, not the competition.
  • The most efficient work teams range from five to 10 in number — a group that can be fed with two extra-large pizzas. The “two-pizza rule” is usually credited to Amazon CEO Jeff Bezos.
  • Another management tip is to distinguish between “Type 1 decisions” that are hard to reverse, and “Type 2 decisions” that are easier to correct if they turn out to be wrong. Much more care has to be taken with Type 1 decisions.
  • When you’re defining the parameters for a major new project, work backward from the press release announcing the project launch. Misener said he has five or six such press releases sitting in a drawer, “one of which I think actually will get out as a press release.”
  • You should be right a lot, but it’s OK to fail. Misener pointed to Amazon’s Fire Phone (2014-2015) as an example. “The hard thing to swallow was that that failure got out,” he said.

Amazon has taken some big bets over the years — for example, when it started offering its Amazon Prime free-shipping plan. “We really didn’t know whether this was going to work,” Misener said. Eventually, the bet paid off handsomely.

“Are we magic? Are we able just to take these big bets that pay off?” Misener asked. “Well, no, we’re not magic. The reason we are able to take big bets is because we’re always betting. … The big mistake is to rely on one thing.”

Who knows? If artificial intelligence can predict whether a tornado will hit two days in advance, maybe it can predict which future business bets will pay off. What was it that Arthur C. Clarke said about any sufficiently advanced technology?

Subscribe to GeekWire's Space & Science weekly newsletter


Job Listings on GeekWork

Executive AssistantRad Power Bikes
Find more jobs on GeekWork. Employers, post a job here.