Sramana Mitra: What about in healthcare?
Paul Daugherty: Healthcare is huge.
Sramana Mitra: How far are we in enabling doctors with AI capabilities?
Paul Daugherty: Just scratching the surface, but there is huge potential. One example people use a lot is machine learning for detecting patterns in radiology and doing diagnostics more effectively. Companies like GE and Philips are advancing those types of platforms pretty rapidly.
Any areas like that where you can use pattern matching technologies are going to be very powerful and will enhance the diagnosis abilities of physicians and other professionals. The other area that’s interesting from an AI perspective is treatment area. There’s an interesting company called NeuroPace that treats epilepsy with an implant that monitors your brain waves in real time.
It understands and learns what kind of patterns in your individual body may be the indicator that a seizure is coming. Then, how can you eject signals to prevent seizures from happening. They’ve already shown that they can reduce seizures for up to 40% using the technology. There’s the learning capability in there to understand and personalize that. That’s one example of many different treatment areas where you can apply AI to change the treatment process.
The other thing that’s really exciting in health is new treatment options enabled by AI. One promising area has been the treatment of people with autism and using AI to help them communicate more effectively. It’s being applied with people who suffer from depression to help them understand the characteristics of the disease and treat it more quickly. In those care type of professions, there’s a lot of potential for AI to understand people and their behavior more precisely that can lead to better health outcomes and treatments.
Sramana Mitra: Going back to my state of the union question, why is it taking so long? Even in the most modern healthcare systems today, doctors take 15 minutes with a patient. They’re expected to do all kinds of past data review and come up with diagnosis and recommendation for preventive care as well as dealing with symptoms. This is a place where AI could work magic.
If all the data and science is programmed into the AI, it could be a tremendous help to make those 15 minute interactions far more powerful. It’s a classic opportunity for augmentation with machines. Why is this not happening yet?
Paul Daugherty: It’s starting to happen. It depends on the country you look at. We have a very complicated health system with very different incentives. You have physicians that provide the healthcare in hospitals. You have insurance as the payers. It’s sometimes difficult to figure out the best way to get innovation into the system. There’s a lot of potential there. Some things are already starting to happen but you’re right.
It would be nice to flip the switch and reimagine the whole system. It’ll take time for it to happen. A particularly impactful area will be something we talk about in our book. We talk about the idea of AI being used to re-humanize time. What we mean by that is, we use AI to offload tasks so people can spend more time with human interaction like the call center example I gave earlier. In healthcare, that’s huge.
Instead of having to access the EMR system or other things that your physician might need to do, what if all they ever had to do is spend time dealing with you and talking to you. Everything else happens seamlessly. That’s the world we’re moving to. It’ll just take a while to get all the technology implemented across the system and change some fundamental ways the system works to allow that to happen.