Skepticism for the MIT crowd

Home - Viewing one post

September 25, 2008

Marketing the book

I turned up the skepticism a tad for a lunchtime talk I gave yesterday at MIT's Tech Review's EmTech conference. I feel a little like a politician saying that. But you figure that people at an MIT conference already know about the Numerati, and understand how dramatically the analysis of massive data is going to change the world. So I could dwell a little more on what could go wrong.

The main point: Most of the systems used to analyse and model human behavior are hand-me-downs from other disciplines. Some of the Numerati (especially at the work place) model us like financial instruments, each person representing a sort of mutual fund of skills, traits, connections. Others use operations research, the same science that optimizes cell phone traffic and flight schedules (and countless other things). Neither one of these approaches can come close to grappling with the complexity of human beings. But if they can figure us out enough to make us a little more efficient or a tad more profitable, they've done their job. The goal of the Numerati, after all, is improvement, not perfection (much less "truth").

So, if we develop and refine these systems, the analysis of humans evolves on analytical methods created for investments and machines. Could that be taking us in the wrong direction? I'd be interested in thoughts. I'm going to be bringing this up at Intel today, in Portland. Maybe they'll have some ideas.

Portlanders: Please come to the event this evening at Powell's!. See you there. Btw, here's Dan Farber's review of The Numerati on CNET. And here's a slideshow I did for Boston.com on 10 Emerging Technologies.