“Got kitesurfing on the mind, mixed with some search & classification tech, and a dab of political ranting”

Is Markoff overstating things a bit?

Posted by direwolff on November 13, 2006

One of Fred Wilson’s post from this weekend titled, Web 3.0 Is The Semantic Web, points to a New York Times article by John Markoff titled, Entrepreneurs See a Web Guided By Common Sense. Which after reading it leads me to the natural next question, “why aren’t technology reporters guided by that same common sense?”. The timeliness of this article couldn’t be more prescient to the rants and discussions I blogged last week on the issue of context. I’m actually surprised that Powerset didn’t get an honorable mention given that two stealth companies and a third who has been around for a long time but with little commercial success, were all mentioned.

Once again, my brainmatter is dwarfed by the notable mentions in Danny Hillis and his start-up, MetaWeb Technologies Inc., and Nova Spivack who is the founder of Radar Networks. Going to each of their web sites will bring that same dissappointment you feel as when you go check out a new movie Web site in November and it only has a splash page telling you the movie won’t be out ’til Summer of the following year with no other details. First off, I have tremendous respect for both Danny and Nova, so I mean no disrespect to them and their endeavors, but while I generally like reading John Markoff’s articles, I’m disappointed when this story, in a space that I’m so close to, comes across like such a puff piece. It’s this kind of story in previous times that made the term “AI” synonymous with unattainable pipe dream as it over promised and under delivered. Thought someone would have learned something from the late ’70s/early ’80s.

Don’t get me started with his first example, “I’m looking for a warm place to vacation and I have a budget of $3,000. Oh, and I have an 11-year-old child.” You know, this is the kind of thing that reminds me of why if a genie gave me 3 wishes for anything I wanted, I wouldn’t use any of them. Try it just for fun. For example what if I wished one of those common good guy wishes like “I wish for world peace”, and the genie turned around and exterminated the earth of all humans including me. “Oh wait, that wasn’t what I meant. I meant world peace for mankind.”, and then all women were exterminated and all men were pumped up with Ritlin. You can keep playing this game and find that for any wish, no matter how noble, there’s always a way to attain it that was not likely your intention, but that is equally effective.

So when I look at Markoff’s example, my first thought is that the system will need to gain some context about who I am and what I would like, not simply be able to do some natural language processing and link that to a bunch of contextually irrelevant information from the Web. Now, I don’t even want to begin addressing issues around how he brings tagging into this discussion, but there’s a good post on a whole other set of issues with this article written by Greg Linden worth reviewing.

I guess one of the things that I find discouraging about this article is that there have been other solutions that have made great strides here, are much less resource intensive, and yield superior results than AI techniques, but because these are not getting the government’s funding attention they remain below the radar. It would have been useful for Markoff to have gone a bit deeper in his exploration of this topic and show some alternatives to solving the problems raised, rather than to rehash what feels like an article from three decades ago or pander to two well known entrepreneurs. Actually, given their mention had to have some PR encouragement, I feel sorry for these two start-ups since they’re already behind the 8-ball on delivering 1/100 of what Markoff has written about. As for Cyc, I’ll let those more familiar with its history raise the sad inconsistencies of this story relative to their accomplishments and promises over the years, but they do remind me of a brilliant line from a Public Enemy song, “don’t, don’t don’t, don’t believe the hype!”.

*** 11/13/06 UPDATE: Dan Farber has a good article discussing the various points of view around this issue and while his discussion is dominated by some of Nova’s comments on the matter, Ross Mayfield’s sobering addition at the end of the piece helps round out the perspectives on this obviously thorny issue of the Semantic Web and Web 3.0.

Tags: , , , , , ,


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: