Saturday, February 19, 2011

IBM Watson makes it official: mankind is Toast

"For one welcome our new feudal lords computer--" that is a quote from Ken Jennings, the guy who used to be egghead "danger" world's largest, until IBM Watson supercomputer waxed the floor with him and his colleague Brad Rutter, puny humans.

He was kidding, but not much. In a piece commissioned by Slate, the winner of 74 lots of consecutive danger (until he met his match in Watson) writes:

[Also on InfoWorld.com: to take a hilarious shenanigans of the tech industry, Subscribe to notes from Robert x. Cringely from underground newsletter.]

Watson seems to represent a giant leap forward in the field of natural language processing--the ability to understand and respond to the UK newspaper, the Ask Jeeves has done (with uneven results) in the dot-com boom. Clues for "danger" cover a open domain of human knowledge--every topic imaginable--and are full of traps for computers: puns, slang, wordplay, oblique allusions. But in a few years, Watson has learned--Yes, learn--to address some of the complexity of a myriad of English. When you see the word "blondie", is very good at knowing if "Risk" means the cookie, the comic or the new wave band.

Yes, of course, it helps that success "danger" is highly dependent on how fast you can press a buzzer with the thumb--and there is no thumb faster than an "electromagnetic solenoid trigged by a shock microsecond-accurate." And Yes, we can take some small solace in the fact that very occasionally things wrong Watson--confusing the city of Chicago and Toronto, for example, "The elements of style" with author Dorothy Parker. (The New Atlantis has a detailed analysis of how lost Watson and why--presumably written by a human being.)

But above all, we are Toast, and not necessarily also catch a supercomputer with 2880 CPU and a 15-terabyte database to spread the butter. As uber geek mathematics that emphasizes Stephen Wolfram, you will make your media search engine.

While competitors "danger" medium--almost all of them already among our species best and the brightest--answer correctly the 60 percent of the time (and Jennings pulls off an impressive 79 percent), Google is not too bad either: 66 percent of the time the correct answer to a question of "danger" can be found in the first search result on every page. Bing was 1 percentage point behind but (perhaps because--ahem--uses the Google search results as "signals" to their algorithms), with a few points behind which to ask.

Wikipedia, created and modified by humans? accuracy of 29 percent. Ouch.

What does this mean for the rest of us? For each Jennings:

IBM ... sees a future in which fields such as medical diagnosis, business analytics and support are automated by software question answering as Watson. Just as factory jobs were eliminated in the twentieth century by new robot of assembly line, Brad and I were workers of knowledge-industry first put out of work for the new generation of machines "thought." "Quiz show contestant" could be the first work in excess of Watson, but I'm sure won't be the last.

IBM is already planning to roll out a cybernetic "physician's Assistant" that will help with diagnosis--think of Dr. Gregory House without the sardonic humour or dependence on valium.

Next in the list of endangered species: professional journalist/blogger. Is already in progress. Companies like demand Media already use algorithms to determine which "stories" garner more eyeballs and advertising revenue. It is a very short step to churn as well as the content.

As someone who has been abused on the stories of science fiction where machines suddenly wake up one day and think, "what we need humans, exactly?" this is not surprising nor welcome news.

Will make the new our life easier in the long run of releasing our gray matter for most creative efforts, automation will free our bodies from manual work? Or are we all just end up working for computerized boss--or, worse, as lackeys to keep machines up and running? (I'm sure at least some of my readers already feel that way.)

The machines are winners. Scratch cards that have already won. We only hope that don't realize for a while.

There is still hope for humanity? Illuminate my weekend with some optimistic words below, or email me: cringe@infoworld.com.

For more analysis and commentary on emerging technologies, visit InfoWorld.com. Story copyright © 2010 InfoWorld Media Group. All rights reserved.

No comments:

Post a Comment