[mathattack]

After companies get to be a certain size it gets tougher and tougher to do basic research. Part ofbitbia corporate culture part is fear of undermining existing products.

The examples where this isn't true (lockheed Martin's skunkworks, Apple) are very rate. More common are large companies shifting from innovation to services (IBM an GE).

This makes sense. Will one company necessarily know how to commercialize a technology? This is one reason the Angel and VC models work. They separate the innovation from big company politics, and create an enviroent where killing old products is encouraged.

[m0th87]

I would argue the exact opposite: larger companies tend to have enough resources floating around to be able to perform research. e.g. Bell Labs, PARC, Watson, BBN...

VCs have their place, but they fund ideas with shorter runways. It's the opposite model to research.

[mathattack]

They have the resources to invest, but struggle to commercialize. This is why so many of Xerox's great inventions were commercialized elsewhere. Even when they have success stories (ibm's first PCs) they are usually hives off from the rest of the firm.

Smaller firms lack development dollars but have more flexibility in finding a market niche. For early stage innovation, being small also can force simplicity.

[digikata]

GE may play in the services game, but I see them as fairly innovative in a niche of complex systems engineering with real hardware in a way that few of the larger companies quite sustain - their jet engines, and wind turbines to name a couple areas.