Saturday, July 24, 2021

The Logic Behind IBM’s Watson Healthcare Exit

I’ve been following and fascinated by IBM’s advanced enterprise artificial intelligence platform named Watson since the very start.

When it started, the very first implementation I was briefed on was for medical work. It showed a ton of progress, and I still recall getting briefed by the leading physician working on the project. He told a story of a patent he’d become interested in that had a disease that no one could diagnose, so he took on the case and, after two years, discovered the cause and prescribed a cure. The disease was both excruciating and very debilitating, and the years that poor woman suffered would have been like torture.

When they put her symptoms into Watson, a prototype at that time, after around 15 minutes, it kicked out a list of possibilities. The obscure disease they had found she had was in the top ten. Watson promised to revolutionize medicine, so why is IBM abandoning it? The reason is likely because regulations are making it impossible for a tool like Watson to work.

Let me explain.

Data Access

IBM has complained about a serious problem ever since they donated massive resources to help find viable Covid-19 Medications. That problem is the inability to access much of the massive amount of existing medical research and patent data. What makes an AI work is data access, and if you have it, you have an intelligent AI. But if you don’t or the data can’t be consistently indexed, you don’t.

Also, there is undoubtedly a lot of bad data in these massive medical repositories, given they don’t have to meet any global standard, and the risk of breaking laws across borders for accessing this data is exceedingly high.

IBM has been working on this data access problem from well before the pandemic hit and has devised tools to deal with the lack of compatibility, but vetting these massive repositories is beyond their capability. However, conflicting regulations have been equally problematic.

But that isn’t even all of the problem. Medical facilities tend to be very possessive of their data. They tend to want to be compensated if that data is used to create something like a medication that can generate revenue. Their sense of value can be far higher than sustainable, but they often make their money by knowing more about curing a disease than any other facility. Allowing that knowledge to flow worldwide, even if it saved millions of lives, is often not something an administrator compensated on profits will allow.

So IBM had to plow through three layers of problems, incompatibilities, regulations, and a broad unwillingness to provide access to critical data needed to make Watson rise to its full capability.

What Is Being Missed

Now imagine being able to fill out a form, run a simple scan at home, and be accurately diagnosed remotely for even the most unknown disease and then receiving not the best resulting solution the hospital offered but the best in the world. Every time. Medical mistakes dropping to a fraction of what they now are, the chance of having an undiagnosed painful illness lasting for years reduced to near zero percent, and far lower chances that you’ll lose a loved one or your own life.

This outcome is the result that Watson could provide, and IBM knows it. But because of problems IBM has no control over, they can’t get to that remedy. Granted, they are spending so much on the effort that it is unprofitable even though it is generating over billion dollars in revenue. Still, the money issue likely pales in comparison to the frustration of having a solution that could not only improve the lives of millions but save the lives of millions, but they can’t get there.

It would be like having your heart’s desire always just out of reach. You can see what you need to get there, the tool works, the result is physically achievable, but the bureaucracy surrounding the data side of the problem is making it impossible to get there.

That’s likely why they want to sell to a company focused on medicine, both because that company may know how to mitigate some of these problems and because watching people get sick and die that you could have saved proved exceedingly depressing.

Wrapping Up

Few things depress me more than seeing bureaucracy get in the way of technology that could make lives better and last longer. IBM has been working for much of this century to bring one of the most potent AIs ever developed to the problem of human health. But they have faced a massive uphill battle at every turn, and they just got tired. It reminds me of the story of the guy hitting his head against the wall being asked why he was doing that? His answer was because it felt so good when he stopped. IBM wants to feel good again, and I can’t blame them.

Here is hoping they find that one magic buyer that can accomplish what they didn’t, bring Watson to its full potential. Frankly, I think this effort needs to be driven by the government, given it is the government that is causing many, if not most, of the problems. If you think about it, correctly implementing it could lower healthcare costs by identifying cheaper remedies that work, removing the need for excessive testing, and eliminating many of the causes for malpractice.

A rough estimate would be a reduction in medical costs by over a third, but, for now, that is the path not being taken, and we’ll suffer for it.

Similar articles

Latest Articles

Data Science Market Trends...

When famed mathematician John W. Tukey postulated that advanced computing would have a profound effect on data analysis, he probably didn’t imagine the full...

Data Recovery Market Trends...

Data recovery is more important than ever in this era of constant cyber attacks and ransomware. The Verizon Data Breach Investigations Report (DBIR) looked...

Trends in Data Visualization

In a world of big data, visualization is becoming a key skill set that every business must master.  Digital technology has transformed the way businesses...

Microsoft Data Portfolio Review

With a host of analytics services for almost any situation, Microsoft Azure’s data services have got just about every base covered.   In the world...