IBM’s Watson is the brilliant machine that crushes Jeopardy champions, analyzes sports statistics to improve player performance, and churns through 60 million pages of medical journals every second to learn how to diagnose diseases and recommend treatments.
Soon, Watson and programs like him may power every business, one expert says.
They’ll be used to predict the most popular cupcake frosting at your local mom-and-pop cupcake shop. Your accountant, your lawyer, even your barber may soon be using artificial intelligence, says David Cox the IBM Director of the MIT-IMB Watson AI Lab, a first-of-its-kind industry-academia collaboration between IBM and MIT. The program focuses on fundamental research in artificial intelligence (AI), which essentially refers to the simulation of human intelligence by machines; namely, advanced computer systems.
AI programs like Watson have advanced considerably and their capabilities are growing by the nanosecond. Cox believes that will soon have major implications for Main Street businesses. His research is focused on bringing insights from neuroscience into machine learning and computer vision research, and his academic lab has already spawned several startups across a range of industries, from healthcare to autonomous vehicles.
We spoke to the IBM thought leader about how entrepreneurs from every industry can harness artificial intelligence to benefit their businesses. Here’s what he said:
- Every business everywhere will eventually depend on AI.
“There is a joke in the VC world that every start up is an AI startup. Because AI is sexy. Even companies that aren’t in the space claim to be AI in order to get funded.
“But that said, pretty soon every company will genuinely be an AI company. AI will be so integrated into everything we do, it will be like software. I don’t care if you run a grocery store or a legal consultancy. It will be woven into everything. Every company at every scale will have to have some strategy to harness it or they risk being left behind.”
- You can get ahead of the learning curve.
“We are in a stage where most of the people who can usefully deploy AI have PhDs and computer science. It’s a very thin slice of the population. And that was also true in the early days of the software age. But access to knowledge is increasingly opening up. There are free classes online at places like Coursera or YouTube where you can really learn a lot.
“There are also a lot of tools in development to help people learn AI technology. There is a learning curve, just like there was during the dawn of software. And just like software there’s a cycle of hype and fear and then everyone will price it in. The faster you learn it, the more of an advantage you have.
- Define the problems you need to solve.
“AI is a tool. But you won’t be able to deploy it correctly if you don’t understand why you are using it. Don’t fall in the trap of just learning the buzzwords. Really try to define the problems you are trying to solve. Right now AI is really good at attacking big data. But most businesses actually have small data that they need to harness, and they will have to learn methods of addressing this. It’s a problem you can only solve by knowing what your problem actually is.
“There’s going to be dramatic progress in the next five years with regards to AI. But knowing the problem you are trying to solve will always be at the core of what someone can get done with AI.”