IBM Watson Health Layoffs Suggests AI Strategy Isn’t Working

Posted on June 6, 2018 I Written By

Anne Zieger is a healthcare journalist who has written about the industry for 30 years. Her work has appeared in all of the leading healthcare industry publications, and she's served as editor in chief of several healthcare B2B sites.

IBM Watson Health is apparently making massive cuts to its staff, in a move suggesting that its healthcare AI isn’t working.

Watson Health leaders have argued that AI (which Watson Health leaders call “cognitive computing”) as the solution to many of the healthcare industry’s problems. IBM pitched Watson technology as a revolutionary tool which could get to the root of difficult medical problems.

Over time, however, it’s begun to look like this wasn’t going to happen, at least for the present. Among other high-profile goofs, IBM Watson has struggled with applying the supercomputing tech to oncology, which was one of its main goals.

Now IBM Watson Health has slashed up to 70% of its staff, according to sources speaking to The Register. The site reports that most of the layoffs are cutting staff within companies IBM has brought in an effort to build out its healthcare credentials. These include medical data company Truven, acquired in 2016 for $2.6 billion, medical imaging firm Merge, bought in 2015 for $1 billion and healthcare management firm Phytel, the site reports.

The cuts reflect a major strategic shift for Watson Health, which was one of IBM’s flagship divisions until recently. Having invested heavily in businesses that might have helped it dominate the health IT world, it now appears to be rethinking it’s all in approach.

That being said, no one has suggested that IBM Watson Health will disappear in a poof of smoke. IBM corporate leaders seem dedicated to an AI future. However, if this report is correct, Watson Health is being reorganized completely. Not too much of a surprise since given how hyped it was, it would have been almost impossible for it to live up to the hype.

To me, this suggests that rolling out healthcare AI tools might call for a completely different business model. Rather than applying brute force supercomputing tools to enterprise healthcare issues, it may be better to build from the ground up.

For example, consider Google’s approach to healthcare AI supercomputing. UK-based DeepMind is building relationships and products from the ground up. Working with the National Health Service DeepMind Health is bringing mobile tools and AI research to hospitals. Its mobile health tools include Streams, a secure mobile phone app which feeds critical medical information to doctors and hospitals.

In my opinion, the future of AI in healthcare will look more like the DeepMind model and less like IBM Watson’s top-down approach. Building out AI-based tools and platforms for physicians and nurses first just makes sense.