IBM Watson is a computer system created to answer questions – questions posed in unstructured, free, natural English language.
It answers using knowledge that it builds for itself by reading and understanding the contents of books and other documents.
It learns how to identify answers by being trained, learning from experience how to interpret the evidence in it’s knowledge.
I wrote a (long and rambling) post in January about work that had been done on Watson since it was unveiled to the world.
Here’s a quick overview of a few things that’s happened since then.
Watson is being used by WellPoint
Watson is now in production at WellPoint (a health insurer in the U.S.). They’re using it to support pre-authorisations for medical procedures. This was a very manual process: authorisations involve teams of trained nurses who review patient cases, sometimes taking weeks to complete.
An instance of Watson built a set of knowledge from a variety of medical sources, and was trained using the experiences from the histories in WellPoint’s previous cases.
WellPoint’s nurses now submit patient cases to this Watson for review.
This should increase the use of evidence in the process as cases can be reviewed using the knowledge from a massive amount of the latest medical literature. It should also make the process much quicker.
WellPoint talked to Fortune Magazine in September about the progress of the pilots:
“…[Watson's] ability to learn… As we’re working with Watson today, the nurses are ranking and scoring the responses — right answer, right reason, outcomes as a result. We continuously feed that information in, and Watson just keeps getting smarter…”
A video from an early pilot of this work was released in May:
Watson is in pilot programs as an oncology diagnosis & treatment advisor
Experts from the Memorial Sloan-Kettering Cancer Center (MSKCC) in New York are helping to teach an instance of Watson about cancers.
This Watson has built a knowledge from sources including MSKCC’s renowned library of cancer research, and been trained to diagnose patients and make treatment recommendations by some of the best cancer doctors in the world.
This is still a pilot program at the moment, so is currently only available to a select group of oncologists. Look out for it being made more widely available to doctors in the future.
Fast Company magazine wrote about this effort in October:
“…Watson is poised to change the way human beings make decisions about medicine, finance, and work…”
A video from an earlier phase of the pilot was released in March:
Research is continuing
While Watson is starting to be put to work in real-world and commercial environments, there is still Research work continuing to make new breakthroughs in the underlying science and technology.
One aspect of this has included collaboration with the Cleveland Clinic, looking at the application of Watson in the medical training field. This will provide medical students with access to the latest journals, articles and medical reference sources, while allowing them to team and train Watson and make it smarter over time.
“…Part of Watson’s training will be to feed it test questions from the United States Medical Licensing Exam, which every human student must pass to become a practicing physician. The benefit for Watson should be to have a difficult but measurable set of questions on which to measure the progress of its machine-learning technology…”
A video announcing this work was released in October:
Watson is being tried in industries beyond healthcare
Watson could be used by call centre agents to quickly get deeply-personalised answers for the client they’re supporting on the phone.
A demonstration of this was given in October:
(skip to 26:35 – unfortunately Livestream’s embed code won’t let me do that for you)
Is that all?
There has been a lot more that’s happened this year, but these are some cool highlights.
I’m trying to get better at this whole “being concise” thing.