I read “Make IBM’s Watson Your Personal Shopping Assistant.” IBM wants to leapfrog www.pricewatch.com, www.amazon.com, and the aging www.mysimon.com, among other shopping services.
Now quite a few people have embraced Amazon’s flawed, yet popular, recommendations service. I am trying to remember when I first noticed this somewhat annoying feature of the digital WalMart. I cannot recall. I am reminded of the weaknesses of the system each time I log in and see recommendations to my wife’s book selections. Undoubtedly she and I are not following Amazon’s best practices. My wife is pretty familiar with my user name and password, Amazon, and the ease with which she can order products (dog vitamins), novels (wonky mysteries infused with herring), and oddments I know won’t plug into my computer systems; for example, something for a faux soft drink machine.
My view is that for some folks, an Amazon habit is going (note the present progressive) to be difficult to modify. Even though Amazon is struggling to deliver profit joy, the Amazon online shopping thing has quite a following.
Well, just in the nick of time–is it years too late?—IBM says it will apply the billion dollar baby to meet my shopping needs. Oh, yeah. Here’s what I learned from the write up:
IBM is partnering with Fluid, a digital commerce company, to create a one-on-one experience with Watson’s capabilities. For example, let’s say you’re looking for the perfect gift for your significant other. Tell Watson about the likes and dislikes of your loved one and let the computer score through piles of data, and eventually pick out a product (or group of products) with those details in mind. Or let’s say you’re going on a hike in the Himalayas and need the right gear for your trip: once you tell Watson what you need, the computer does the research and picks out all the right equipment for you.
I suppose this means that Amazon’s reviews are about to be staring at Watson’s tail lights. The article doesn’t pay much attention to Amazon or lesser services that pepper Google results pages with offers of prices, reviews, and suggestions for the procrastinating Mother’s Day shopper.
IBM is working on an app for XPS that will work on desktops, tablets and smartphones. It will be able to ask the same sort of questions you’d expect from a salesperson in a physical store, but without the hard-sell techniques and with a lot more personalization.
I think my grade school teachers called this the present progressive. I translated this to “it may sound now but nothing is showing up right now.”
First, is IBM or a “partner” going to design, build, debug, deliver, and support this magic carpet shopping service? On one hand, it looks like Watson’s brain trust in Manhattan is on the job. Then it struck me that an outfit called Fluid will have to lift that barrel and tote that bale. My hunch is that IBM will watch from the veranda of the hotel overlooking the laborers unloading the good ship Watson.
Second, I keep reminding myself that IBM has yet to provide a demonstration of Watson that makes it possible for me to compare throughput, precision, and recall with the search systems to which I have access. Talk, it appears, is much easier than making and selling a product.
Third, what about that Amazon thing? The Bezos-A-Rama is busy creating yet another digital monopoly. In addition, that big store offers recommendations along with one click shopping, reviews, a so so search system, and fawning Wall Street believers.
To me it looks as if IBM, on the other hand, is doing what IBM does best: Working its public relations firms extra hard. I hear the faint sound of Tennessee Ernie Ford singing,
You load sixteen tons, what do you get
Another day older and deeper in debt
Saint Peter don’t you call me ’cause I can’t go
I owe my soul to the company store
Mr. Ford’s backup singers are IBM’s sales and marketing team after a tough day of talking about what Watson will someday soon be. Hard work is moving 16 tons of marketing.
Stephen E Arnold, April 26, 2014
Phi Beta Iota: All of the “great” companies, from IBM to Oracle to Microsoft to Google, have gone down various rabbit holes and forgotten — if they ever understood — the Holy Grail of computer-assisted intelligence (decision-support): SENSE-MAKING. As a gross general statement, specific instances of success notwithstanding, none of these companies has addressed any of the elements that would comprise a world brain: an autonomous (uncorruptible uninterruptible) Internet with petabyte throughput; massive simultaneous multi-disciplinary multi-domain pattern analysis and anomaly detection, or human-centric sense-making replete with cultural, historial, and linguistic contexts. There is an opening for any company that can attack all three of these problems together, with infinitely scalable open source software as the hub and a range of partners including nano-satellites, humanitarian multi-lingual diasporas, and so on. The industrial “thinking” right now is pedestrian to the point of national embarrassment. Other countries are not doing well either, in part because their super computing is also skewed by their respective secret worlds, and they in turn are peverted by NSA subsidies for doing more of the wrong things righter instead of the right thng (Norway is the latest to fall). The next break out in super-computing will be open, citizen-centric, relatively inexpensive, and will catch the secret world completely by surprise.