Scroll Top

Breaking through the Hype: Artificial Intelligence is the Tool, Not the Final Answer

Once the domain of computer science alone, artificial intelligence (AI) has been hailed as the newest answer to numerous problems in healthcare and the life sciences [1-3]. It seems that everyone is talking about AI without a clear understanding of what it means, what it is good for, and perhaps most importantly, what its limitations are. For every article that touts IBM’s Watson cognitive computing platform’s success in diagnosing a patient with an ultra-rare form of cancer [4] or DeepMind’s ability to match human doctors at identifying eye diseases [5], there are countless others written about an overall lack of substantial evidence that the technology is ready for prime-time [6,7].

One of the most basic challenges to AI (including for people who think and write about it) is coming up with a clear definition of what AI is and isn’t. It’s difficult to say whether or not AI is fulfilling any expectations, if you can’t agree on what it is in the first place [8]! Even when a definition is agreed upon, or a set of definitions (like those developed by Northwestern’s Professor of Computer Science and Chief Scientist at Narrative Science Kris Hammond’s Periodic Table of AI) [9], experts still disagree on the utility of AI currently for healthcare and the life sciences industries. But the key thing these experts are getting wrong is thinking about AI as the answerto healthcare’s problems and not the toolwe can use to get to the answers.

This is an important distinction; consider the conversation around the personal computer in the 1980s. The computer was a tool we could use to make writing more efficient and to perform complex calculations accurately and quickly. The Internet is another example. AI should be thought of similarly, because that way of thinking frees up the industry to pivot to new technologies as they become available. In a recent interview with Wout Brusselaers, CEO and founder of, he described the use of AI and machine learning to accelerate clinical trials recruitment, but noted that at some point, AI will become ubiquitous—it’s not the AI that’s so important, it’s how it’s used and the technology that is built using AI platforms.

In this special MoneyBall Medicine Podcast Seriesfrom the AI Applications Summit, I speak with Wout Brusselaers, Andrew Radin, co-founder and CEO of twoXAR, Ron Alfa, Vice President of Discovery and Product at Recursion Pharma, Guido Lanza, President and CEO of Numerate,Shrujal Baxi, medical director of Flatiron Health, and Milind Kamkolkar, Chief Data Officer at Sanofi. We discuss a variety of topics, from finding talent to new business models and their predictions for where AI is leading us.

  1. Cohen, J. K. 2018.Report: 3 reasons AI in healthcare is no longer just ‘hype’. Becker’s Health IT & CIO Report.
  2. Caban, T. Z. C., Kevin; Painter, Michael; Dymek, Chris;. 2018Hype to Reality: How Artificial Intelligence (AI) Can Transform Health and Healthcare.
  3. Griffiths, S. 2018.Hype vs. reality in health care AI: Real-world approaches that are working today. MedCityNews.
  4. Rohaidi, N. 2016.IBM’s Watson Detected Rare Leukemia In Just 10 Minutes. Asian Scientist.
  5. Summers, N. 2018.DeepMind AI matches health experts at spotting eye diseases. Engadget.
  6. Ziegler, A. 2018.Problems We Need to Address Before Healthcare AI Becomes a ThingHospital EMR & EHR.
  7. Garvin, E. 2017.Don’t Let the Hype Fool You—AI in Healthcare Is Far From Perfect. HIT Consultant.
  8. Marr, B. 2018.The Key Definitions of Artificial Intelligence (AI) That Explain Its Importance. Forbes.
  9. Hammond, K. 2017 The Periodic Table of AI.