Gilster: IBM's Watson leads way into era of machines that learn

CorrespondentNovember 17, 2013 

A San Francisco company named Vicarious claims to have created software that can crack CAPTCHAs, the distorted blocks of text that humans are asked to interpret to prove they are not machines. CAPTCHA stands for Completely Automated Public Turing test to tell Computers and Humans Apart, a system that sprang up in the 1990s to make sure that people signing up on websites were not spambots trying to game the system so they could distribute spam.

This is a big deal if true, for Google’s reCAPTCHA system is considered all but impenetrable to a machine. If the Vicarious software really can, as the company says, solve CAPTCHAs 90 percent of the time, it would be an intriguing advance in artificial intelligence. The company sees it as a first step toward a machine that works like a human brain, and is likening it to IBM’s Watson, famous for its ability to answer the questions posed to it on “Jeopardy!”

The Watson supercomputer was a terrific TV personality, but plans for its future go beyond even Alex Trebek’s wildest dreams. For what Watson is good at is drawing together massive amounts of data to reach a conclusion. A case in point: IBM is working with the University of Texas’ Anderson Cancer Center to find a cure for leukemia, a mammoth task that involves correlating mountains of data, and one that may be aided by the fact that Watson 2.0 is now on the horizon.

Humanlike reasoning

The next Watson is being tuned to scan pictures and interpret them – that would include, for example, X-rays – while down the road we see the prospect of a Watson 3.0, which will move much closer to humanlike reasoning. Fascinating in itself, where Watson may have the most public impact in the short term is in its uses in everyday computing. A first step in that direction is an “Ask Watson” customer service, as IBM collaborates with various companies on the idea.

Taking what IBM calls “cognitive computing” to the consumer has to involve putting it into our pockets. The collaboration with companies like ANZ Bank, Celcom, Royal Bank of Canada and Nielsen points toward eventual smartphone apps that can answer questions, with voice recognition surely coming into the mix. I can only guess what Apple, with its Siri product, or Google may be thinking right now as they ponder the future of talking to your smartphone.

Science fictional device

I’ve found Siri helpful. If I’m in the middle of a project and suddenly remember the need to take a note, I’ll usually just pick up my iPhone and tell Siri to take it for me. Usually the software can come close enough to figuring out what I’m saying that I can make sense of the result, and the quick voice-activated note doesn’t distract all that much from what I’m doing otherwise.

A future smartphone-enabled Watson taken to its ultimate conclusion would be like a science fictional device that could answer almost anything you threw at it. Hold up your phone to an object to get Watson to identify it and give you its history. Ask Watson to diagnose a common ailment by describing your symptoms. Get immediate advice on which car to buy by asking for Watson’s analysis of price and performance. The point is, whenever your decision has to juggle large amounts of data, a device like this could do it faster than your own research.

We’re not there yet, to be sure, but we’re moving into the era of machines that learn, and quickly finding out that high-end data crunchers like Watson can help us make sense out of a bewildering variety of inputs. As we wait for Watson in our pockets, we can hope for real progress in medicine and scientific research through this supple, emerging technology.

Paul A. Gilster is the author of several books on technology. Reach him at

News & Observer is pleased to provide this opportunity to share information, experiences and observations about what's in the news. Some of the comments may be reprinted elsewhere in the site or in the newspaper. We encourage lively, open debate on the issues of the day, and ask that you refrain from profanity, hate speech, personal comments and remarks that are off point. Thank you for taking the time to offer your thoughts.

Commenting FAQs | Terms of Service