IBM not finding it elementary to make money from Watson

IBM (IBM) CEO Virginia "Ginni" Rometty reportedly hopes that the company's "Jeopardy" winning Watson supercomputer will generate $1B by 2018 and $10B within ten years.

The problem is that Watson had brought in less than $100M as of October, as various projects have gone awry or been far more difficult than expected. A service with Citigroup that would recommend financial products to consumers still hasn't been launched.

One great hope is healthcare, with IBM developing a version of Watson that can match cancer patients to clinical drug trials. Meanwhile, insurer WellPoint uses the machine to ensure that treatments meet company guidelines and a patient's insurance policy.

From other sites
Comments (8)
  • user_035922
    , contributor
    Comments (69) | Send Message
    It's one thing to design and build this technology and one cannot argue that it is really impressive stuff. It's an entirely different proposition to invest in training for a Watson-specific team of your employees to help customers implement this solution. Since IBM stopped investing in their employee's training years ago, instead moving down the food chain for sheer cost reduction alone in the form of cheap offshore labor, they are most likely in the position of saying to their potential customers 'we have this Watson thing, but we don't have any employees that understand it and can help you implement it in a useful way.'
    8 Jan 2014, 08:10 AM Reply Like
  • honeycakes
    , contributor
    Comments (5) | Send Message
    They could offer Watson the opportunity to go live near the Plastific Ocean,
    have his own air-conditioned 'mansion' in Mountain View.


    There he'd do some billion-a-minute high-frequency trading
    within the best office hours for time-zone arbitrage.
    8 Jan 2014, 01:53 PM Reply Like
  • Deja Vu
    , contributor
    Comments (1828) | Send Message
    " WellPoint uses the machine to ensure that treatments meet company guidelines and a patient's insurance policy"


    Translation - Wellpoint uses the machine to deny bills submitted by oncology practices. Paying a human nurse didn't work out so well since the nurse could not challenge the doctors, so Wellpoint moved on to using human oncologists. Unfortunately the human oncologists time cost almost as much as the time of the doctor they were trying to stiff. Watson is the latest attempt. Draw up a ever increasing complex web of rules and deny reimbursement. Also claim objectivity by saying we are not doing this, our system is doing this. Soon, very soon, oncologists will stop treating Wellpoint patients or demand payment in advance. Patient will have to brown bag the drugs and pay for the infusion in cash and submit claims to Wellpoint themselves.


    There are reasons why many people root for Obamacare to drive insurance companies extinct. What Wellpoint is doing with Watson to stiff medical practices is one of them. And when bills are denied, guess who ends up in collections and who is angry at the insurance companies?
    8 Jan 2014, 08:35 AM Reply Like
  • User 1538861
    , contributor
    Comments (2) | Send Message
    Watson was built for a specific application - winning at Jeopardy. It's a mistake to view Watson as a solution to general problems. The effort required to apply Watson to other problems (e.g. medical diagnosis, investment advice) is as great as it was building and programming Watson for Jeopardy. Watson is an existence proof that it's possible to use computers for natural language applications - it is not a product ready for general implementation.
    8 Jan 2014, 10:34 AM Reply Like
  • Deja Vu
    , contributor
    Comments (1828) | Send Message
    Totally agree! Searching huge flatfiles for strings and matching them to other strings is not really applicable to every situation.
    8 Jan 2014, 11:03 AM Reply Like
  • 14099282
    , contributor
    Comments (31) | Send Message
    IBM went all in on Watson, they took top notch, front-end & back-end developers to work on it.


    Meanwhile, the rest of IBM (software) is now being outsourced by not only inferior developers, but the work is sub-par and customer base is eroding.


    More buybacks anyone?
    8 Jan 2014, 12:46 PM Reply Like
  • User 20321351
    , contributor
    Comments (232) | Send Message
    * Watson for "Jeopardy" resulted in the R&D of several technologies besides the 'Jeopardy'-tailored last-level layer. They have a lot of potential, and they are already being used in many IBM products. However, I do think that it will take a while to engage big customers, but it will do as it did in the past. Like the 'cognitive' computing effort, they are opening new business fields... which is what IBM always did. They do something really new, stay for a while, and once it is commodity, they sell that part of the business and they start again.


    * IBM still invests in employee training, although it would be good to know how much other companies invest on internal training to have a fair comparison. I do not think that IBM invests much less than other business-related companies. Any pointer to such data?
    8 Jan 2014, 01:01 PM Reply Like
  • Momintn
    , contributor
    Comments (6116) | Send Message
    There are at least 3 apps available that use Watson: Fluid Expert Personal Shopper, Hippocrates, CaféWell Concierge, all powered by Watson.
    Researchers at Cleveland Clinic are using WatsonPaths to support clinical reasoning and are unlocking the promise of electronic medical records with Watson EMR Assistant.
    Houston cancer center MD Anderson taps IBM's powerful Watson computer system to help cure eight kinds of cancers as part of its "Moon Shots" program. The first target: leukemia.
    8 Jan 2014, 06:32 PM Reply Like
DJIA (DIA) S&P 500 (SPY)
ETF Screener: Search and filter by asset class, strategy, theme, performance, yield, and much more
ETF Performance: View ETF performance across key asset classes and investing themes
ETF Investing Guide: Learn how to build and manage a well-diversified, low cost ETF portfolio
ETF Selector: An explanation of how to select and use ETFs