I could be mistaken because it’s been a while, but I read that Watson’s diagnostic capabilities turned out to be mostly marketing and that eventually IBM ended up hiring teams of doctors to process the diagnosis requests that were coming into Watson.
Watson became a marketing term after the company spent hundreds of millions to brand Watson to be synonymous with AI. The term Watson then got appended to existing businesses as it allowed them all to benefit from the brand equity and Watson ads. This unfortunately happened even if there wasn’t any AI capabilities, so it eventually backfired.
Watson Health seems to have been focused on selling the narrative of AI in healthcare, even though the technology wasn’t there.
The divestiture is only for IP also, and it seems most people in the group will be laid off.
As someone with no inside knowledge, it seemed to me that watson started as a technology (or maybe solution/set of solutions) and as time went on, it was pivoted to be a brand? Hard to tell for sure with how difficult it is to get IBM to answer questions about what they actually do...
Worked for IBM for three years, this is accurate. To solve some clients problem we would build an ML solution from scratch just like everyone else, and then try to shoehorn some Watson service into it so we could use the Watson Brand to distinguish our product.
The solutions we built were generally pretty good and our clients were happy, but the Watson part was never anything more than marketing,
Sounds like what happened at Theranos! I read the analysis Watson was generating was ultimately just ignored by doctors because it came to inaccurate conclusions, so that makes sense.
I personally feel Watson was an extremely clever marketing boondoggle. If you think of it, machine learning, neural networks and AI were just making a return into the public mind around the time they announced Watson.
I think somebody thought if they "humanized" AI by making it seem like it was a character, it would make AI seem all that much more closer to the dream.
On the face of it, not a horrible idea, but applied to what was essentially a bunch of separate algorithms.. pretty misleading, but that's just an opinion.
Worth noting that Jeopardy Watson had very little ML and absolutely no deep learning (it was a few years pre-AlexNet). I don’t even think it used any neural networks; certainly not in any major way because they’re not a major topic anywhere in the press releases, working group notes, or the papers published by the Watson research group. Watson was an incredibly complex mixture of bespoke implementations of “classical” AI and NLP techniques to handle questions of different classes by transforming them into search & information retrieval problems. They were able to make it work pretty well for the very limited domain of questions that arise in Jeopardy, but it was also obviously a Herculean task to generalize that approach. I can totally believe that as executives started to grok what Watson really was they realized that it had more value as a brand than as a technology.
Like in Neal Stephenson's Diamond Age with the on-demand voice actors.
I wonder if that was a lack of imagination about what AI will become, or if it was prescience that AI will turn out to be cost ineffective or not capable for most uses.
