At one point IBM Watson Health featured articles that claimed it might cure cancer. But IBM’s flashy coverage of its AI brand targeting the healthcare industry may have been an example when the hype around a particular technology – AI – trumped that technology’s actual capabilities.
After a few high-profile public failures over the past several years, IBM announced it was selling parts of its Watson Health business to private equity firm Francisco Partners — a sale many in the industry had been anticipating over the past year.
Assets sold include data sets and products from the many acquisitions IBM completed to move into the Watson Health brand including Health Insights, MarketScan, clinical development, social software management, Micromedex and imaging software. In announcing the deal, the companies said Francisco Partners would hire key members of the Watson Health team and build its own businesses in the future. IBM will retain its Watson brand, which includes many other vertical industry offerings, as well as continue to support existing healthcare customers and partners with many other products.
Watson cures cancer
Get out of danger! The wins of 2011—whether in the game itself or in turning the company’s software into something of a celebrity—maybe Watson seemed poised to solve the world’s biggest problems, like predicting weather and curing cancer.
“This is just the story of the hype cycle playing out again,” says Jeff Krebs, senior vice president of research for Gartner. “What happened was that marketing was way ahead of research and engineering, and that ultimately caught up with implementation.”
Right out of the gate, IBM selected a type of Watson Health issue – cancer – in partnership with lead client and reference partner MD Anderson Cancer Center to use Watson Health to accelerate clinical decision-making and align patients with clinical trials.
Can Watson win the support of oncologists?
There are some tasks in healthcare that AI performs well – for example, image classification and natural language processing. But the project called on Watson Health to perform tasks such as supporting a treatment decision, Krebs says.
“These are not just perceptual tasks, these are executional tasks that require a lot of steps and rules based between the perceptive parts,” Krebs says. “This has been a huge area for dealing with the state of artificial intelligence today.”
Paddy Padmanaban agrees. He is the founder and CEO of Damo Consulting, a growth strategy and digital transformation consulting firm that works with healthcare organizations and other technology companies.
“One of the challenges in oncology is being able to come up with insights that an oncologist will find useful or valuable,” he says. “Cancer care is very complex, and human judgment and intuition are very important. If technology cannot enable humans to perform better, or if recommendations don’t seem right, oncologists will not use it. If they don’t use it, it will be a downward spiral for technology.”
Padmanaban says the problems MD Anderson has had with Watson Health have been very public, and this may have damaged the reputation of the Watson Health brand.
Negative press from MD Anderson’s relationship raised a lot of questions in people’s minds about whether they should use Watson Health because they “don’t seem to be having a good time at MD Anderson.” “Maybe the technology is not ready? Not to mention all the costs. Do we know what that will achieve?”
Padmanabhan compares this healthcare solution and partnership with other technology/healthcare partnerships like Google and Mayo Clinic, announced in September 2019. The strategic partnership combines Google’s cloud capabilities and artificial intelligence with Mayo’s clinical expertise to improve the health of people and communities by understanding insights More widely, according to a blog announcing the deal. It looks like another shot of the moon, but behind it five more years of research and development.
This month Microsoft announced the Artificial Intelligence Innovation Alliance, which brings together several healthcare and academic organizations to “promote health by identifying and addressing significant societal and industry barriers,” according to the announcement.
Healthcare is clearly an area where the big tech companies believe AI can make a difference. Just how that will look in the future is a different question.
AI for healthcare becomes practical
Krebs says the use cases he sees for AI in enterprise organizations, including healthcare, have become more practical in the past few years.
“The healthcare AI market is moving quickly from vision to implementation,” Krebs says. “Three or four years ago, it was very easy to walk into the healthcare executive board room and really wow people with some heavenly thinking — innovative use cases, new approaches to new data sources — the big data gun.”
But the market is different today.
“This is not the buying market or the need for the market today,” he adds. “Today is really about what we can achieve in the near term by using this new tool.”
What to read next: