[PaulHoule]

Well, the most innovative company of all time was AT&T back in the monopoly days.

They gave us the transistor, Unix, and the C programming language. They hired people like Claude Shannon and Dennis Richie.

Then they broke up AT&T and now "innovation" at AT&T is peddling an inferior DSL service and trying to figure out how to get you to pay more for your cell phone bill.

YCombinator-style startups might make something like AirBNB, but they won't make something as transformative as the transistor.

As for lean and agile, these are both profoundly conservative forces, in fact, that's really their strength. At their best, they improve the odds of project success by squeezing out risk. However, they also encourage people to keep their heads down and not ask the kind of fundamental questions that can lead to 10x gains rather than 10% gains.

[jaydub]

The article makes a point that impactful change tends to come from trial-and-error over a longer period of time. I think in startups it's possible but its a lot harsher environment where only the strong/successful ones survive early to mature and spread their innovation. Maybe at a larger company the same experiment would have had more time to incubate, perhaps because (at e.g. Bell Labs) the parent could afford to invest in basic research for a long time. Or maybe a startup founder would see a problem in a very different light and attack it in a way that a big company (perhaps AT&T today) would never OK in the first place.