IBM Watson Crushes Humans. Is Google Next?
February 16, 2011
The Web is abuzz with IBM Watson’s victory over humans on the TV game show Jeopardy. In my father’s independent living facility, a number of ageing humans expressed concern that their medical diagnoses would be handled by the human crushing Watson.
Here in the airport, the response to Watson’s victory over humans seems more subdued. I asked one college student sitting next to me if he were worried about IBM’s supercomputer taking over the world. He said, “Eh, what, dude?”
National US media were excited about IBM’s triumph on a TV game show. In “Beyond ‘Jeopardy’: Watson Wins,” MSNBC said:
Watson was built to serve up quiz-show knowledge, but those question-answering capabilities would probably be most valuable in specialized fields such as medicine and law. Watson’s kin could help us puny humans sift through millions of possibilities and come up with the five or six best medical diagnoses, or legal precedents, or chemical configurations, or … well, you name it.
Okay, I will name it: PR.
In a clever attempt to regain the technology champion award, IBM—a $100 billion enterprise—showed its information technology on a TV game show. How appropriate for a company that created STAIRS III, Juru, Web Fountain, and dozens of other search and content processing systems. Most recently, IBM was groping for a solution to clustering in its ageing document management system. Its “flagship” enterprise search system is essentially the open source Lucene system.
My question is, “If IBM’s information retrieval technology is so darned good, what’s up with the clustering issue in its records management system?” And “Why is the IBM enterprise search solution based on Lucene?” And, “What has happened to STAIRS, Juru, et al?”
My view is that IBM does not seem to have much traction in the commercial, enterprise search space with its own technology. The IBM demo approach is marketing, and I think it is great public relations.
But where it counts, IBM is far behind Autonomy, Endeca, Exalead, and dozens of other vendors with enterprise solutions that work and are affordable.
What about Google? Does Watson frighten Messrs. Brin and Page? Are the wizards at Microsoft Oslo shaking in their ski boots?
I don’t think so.
In my opinion, IBM is far behind in search and content processing technology. IBM resells other vendors information technology, acting, it seems to me, more like a consultant than an innovator in information retrieval. After buying Cognos and SPSS to bolster the firm’s data mining business, IBM is going to have to do more than beat meat on feet on a TV game show. IBM now has to win head to head procurements for search in the enterprise. Do you think that might be more difficult than winning a TV game show that is popular among those with whom my father hangs out, napping during commercials and shouting questions to the host’s answers?
I do.
Just my opinion. Honk.
Stephen E Arnold, February 16, 2011
Freebie unlike the costs of producing a TV game show, editing the program, and pumping hyperbole into the info stream.
Comments
2 Responses to “IBM Watson Crushes Humans. Is Google Next?”
-
- Subscribe to Beyond Search
Feature archive
News archive
-
[…] This post was mentioned on Twitter by Stephen E Arnold, singularitytrap. singularitytrap said: IBM Watson Crushes Humans. Is Google Next? http://chtr.it/JLq48u #singularity […]
What IBM’s academic paper on Watson said about Lucene and search techniques:
http://www.stanford.edu/class/cs124/AIMagzine-DeepQA.pdf page 71:
A variety of search techniques are used, including the use of multiple text search engines with different underlying approaches (for example, Indri and Lucene), document search as well as passage search, knowledge base search using SPARQL on triple stores, the generation of multiple search queries for a single question, and back?lling hit lists to satisfy key constraints identi?ed in the question.