the Onda

Everyone is aflutter today about the fact that Amazon's web services, including its indispensable Simple Storage Service, took one on the chin with a multi-hour outage. Apparently the entire productivity-sucking swath of the Web 2.0 economy sputtered along with the outage (no profile pictures on Twitter— OMG!!) and folks started screaming bloody murder about the coming apocalypse.

Interesting that one of the first to report was Nick Carr, whose new book "The Big Switch" I just finished reading. The book's central argument is that there is a platform shift coming in computing that will be very similar to what happened with electrification, with computation and storage moving from a distributed model (loads of server racks managed by IT) to a centralized one (a la Amazon Web services) where a few vendors will run huge compute clouds that the rest of us will plug our applications/appliances into.

The book makes a great argument, though it could just have easily have been made in 50 pages instead of 233 (including all of the fab historical color around what happened during the big switch of electric power generation)— an argument which is no less compelling because of today's outage. The reality is that these types of glitches happen at all scales, and though they hurt a little more when its a big centralized provider (witness RIM's second big outage the other day), some economic forces are just too strong to fight.

But what the book doesn't cover well is the way in which opportunities for innovation move around when these big shifts happen. I was really looking forward to understanding how centralized, ubiquitous, and cheap electricity made it possible for a host of appliance vendors to invent whole new product categories around the home and office. And though Carr touches on just this, the extension of the analogy to the current computing shift falls flat with a lame discussion of consumer mashups, a brief tour of personal publishing tools available today, and some recycled speculation on how Google really is trying to build an omniscient AI.

I would have preferred a more nuanced parallel between the first mass-market electric gizmos and the new appliances of the centralized computing grid (i.e., the iPhone, the Tivo, the Chumby, even the bulk of today's laptops). After all, it is only when we start to think about how these new appliances that consume compute cycles from the cloud will change our lives in a permanent way that we can really pause and take measure of what outages like today's will really mean 2,5, and 10 years from now.

Perhaps that is coming in his next book...

Post a comment