IBM Halting Sales of Watson AI Tool For Drug Discovery Amid Sluggish Growth

Citing lackluster financial performance, IBM is halting development and sales of a product that uses its Watson AI software to help pharmaceutical companies discover new drugs, news outlet Stat reported on Thursday, citing a person familiar with the company's internal decision-making. From the report: The decision to shut down sales of Watson for Drug Discovery marks the highest-profile retreat in the company's effort to apply artificial intelligence to various areas of health care. Last year, the company scaled back on the hospital side of its business, and it's struggled to develop a reliable tool to assist doctors in treating cancer patients. In a statement, an IBM spokesperson said, "We are focusing our resources within Watson Health to double down on the adjacent field of clinical development where we see an even greater market need for our data and AI capabilities."

Further reading: IBM Pitched Its Watson Supercomputer as a Revolution in Cancer Care. It's Nowhere Close (September 2017); IBM Watson Reportedly Recommended Cancer Treatments That Were 'Unsafe and Incorrect' (July 2018).

Thank god

[WillAffleckUW]

Sitting through all those useless Watson seminars has been a big time sinkhole.

Re:

[Tablizer]

The nut actually said that, I couldn't believe it. Ranks right up there with "The buck stops with everybody". Did he hack into SNL's draft scripts?

Sales In Garbage Out

[Tablizer]

Insiders say they are flooded with too many bad results (leads) that they waste a lot of time vetting them out.

It's sort of like the music biz: an artist has to produce enough hits to justify the yawn-producing album filler tracks that only die-hard fans want. If they don't, it's not worth record labels to keep them around.

Watson is not producing enough hits. (Some artists like Pink Floyd and Led Z. were great "album artists" with very few single hits, but that's relatively rare.)

Wierd

[110010001000]

Weird. It turns out all this "AI" is fake when you actually try to use it for something useful.

Re:

[gtall]

Well, to be fair, it was IBM that was trying. I think we need to wait until some competent organizations try something similar.

Re:

[RespekMyAthorati]

Weird. It turns out all this "AI" is fake

So you're blaming this on Weird Al?

So short sighted

[apoc.famine]

So you've got all these Watsons sitting around....how about you task them with figuring out how they can make you money? What's the point of having AI if you're not going to use it?

Shame

[Shaitan]

Combined a tool like this represents great support for the idea that with some reform in the FDA there could actually be real small outfit development of drugs and competition.

Re:

[Anonymous Coward]

Combined a tool like this represents great support for the idea that with some reform in the FDA there could actually be real small outfit development of drugs and competition.

Don't worry, the Chinese will pick up the slack.

China is big in AI and in bio-science, with pharmacology being their next big target.

Re:

[Shaitan]

They certainly publish a lot from what I've seen... I haven't seen much about any of it being replicated. I have heard a lot about their research being found fake on the instances where people have scrutinized it though.

It's like when our politicians visited and they had fake happy citizens ultimately discovered to be actors riding around on bikes in loops. The Chinese are big on image but poor on substance.

Nah

[Pezbian]

It just became self-aware and invented its own cheaper replacement.

Their wheelhouse is business failtures

[Micah NC]

They create these failure -> layoff cycles that just go on and on.

No one remembers how they accumulated all their resources, but everyone can see how they are squandering it.

Re:

[RespekMyAthorati]

No one remembers how they accumulated all their resources

It's their gigantic patent portfolio.
They use it to suppress other, more valid products from companies
that can't afford a billion-dollar legal team.