I am also old and have been around the block. There's a chance I hold the record for having the most kids of any founder in a YC batch, which I think is astonishingly cool...
I think it's rare that startups have anything as ground breaking as page rank, and even rarer that it seems ground breaking to anyone else at the time. If we had some specific formula for growing environment saving bacteria from leftover banana peels, we'd tell you for sure.
What we really have is a set of guiding principles, indicators that tell us we're pointed roughly where we should be pointed, and lots of things we want to try to get there.
The tech details weren't meant to be evasive as much as de-emphasized. They're there because we think good candidates might find those specific technologies interesting. We had variants on the description with many more specifics, but none of them seemed entirely important at the time. I do like talking tech, though, so here's an exhaustive look:
* We use CoffeeScript at several levels and happen to think it's pretty awesome. Anything that we could have done in JS has been done in CoffeeScript instead, this includes super tight ad serving javascript, the frontend code for our user facing apps, etc.
* We have two major classes of users, those who use our tools to do better advertising, and those who see ads. The "magic" for the first class is whatever the latest release candidate of Rails 3.1 is (rc5 maybe?), the second is the Node.js version that Heroku's Cedar stack supports (0.4.6 I think).
* Both apps feed copious amounts of information into a backend Mongo replica set, and the Node layer leans on Redis for some things.
* We've done limited data visualization with that d3 toolkit I mentioned. We feel that data viz is often underrated, but an extremely useful thing for us to devote resources to. If we could justify buying the entire New York Times data visualization team we'd do so.