What it is: Pundits, start-ups, and marketing firms right now are throwing the term “Artificial Intelligence” like confetti on a parade. It’s the new “greenwashing.”
What makes an AI system different from a really good program with massive computing power and a monster database is that the computer will only do what you program it to do, with the information you give it. AI learns from experience, and can do things you don’t expect for reasons you can’t fathom — hopefully with better results than you would get on your own. Put it to work on a thousand cases of pancreatic cancer and it will learn what works and what doesn’t and bring recommended solutions that you might not have thought of.
Stanford/Duke/Singularity University futurist and general really smart guy Vivek Wadhwa has an op-ed piece in the Washington Post about how AI has really picked up speed and is about to land in our laps — and everywhere else in our lives. As he says, “There have been more advances in AI over the past three years than there were in the previous three decades.”
Why it’s important: Wadhwa quotes the inimitable Kevin Kelly (author of the new The Inevitable: Understanding the 12 Technological Forces That Will Shape Our Future, cofounder of my alma mater Wired Magazine, and somebody you really want to follow if you’re interested in the future) as saying that AI today is like electricity 100 years or so ago, a utility that will infuse everything we do: “There is almost nothing we can think of that cannot be made new, different, or interesting by infusing it with some extra IQ. In fact, the business plans of the next 10,000 start-ups are easy to forecast: Take X and add AI. This is a big deal, and now it’s here.”
Will it work? Absolutely. I would bet 2,000 of those 10,000 startups are in healthcare. There is no field I can think of that would benefit more from an “add AI” mindset.
But — and this is a big caveat — you can’t just slather on AI and expect it to make healthcare work better. If we apply AI to healthcare with the same lunk-headed inattention that we have used in building many of our present information systems, it will have the same result but on steroids, further clogging the workflow, overburdening the clinicians, and leading to big clinical mistakes.
Remember that Trust is one the seven levers of change in healthcare. In information systems, trust is the quality that allows unimpeded information flow between nodes. Between humans, trust allows the unimpeded flow of information, judgment, and execution. For AI to improve and not degrade clinical workflow, it has to be built with the specific goal of both enhancing and streamlining the clinician’s flow of trusted and trustable information and judgment.