 Silicon Valley and Route 128 and later on Gilson, as long as the probability for any given innovation does not decline more rapidly from the well-behaved innovations, if you're basically imagining that the cost of starting a business is on the access and we're seeing it with red, once you reach a certain point, you can have many, many, many, many low-level innovation. Even if the ones at the higher end of the distribution are more carefully planned and more justified because they were able to be capitalized, they have to be sufficiently more likely to actually succeed than all of these small efforts here, both market and non-market, in order at a system level to perform better. That doesn't seem to be what we're seeing, which is why I think we're seeing this very distributed model generating innovations from where we didn't expect them. Institutionally, that means that freedom to operate is more important than power to appropriate. And this, I think, is the core institutional point about openness. That means commons or symmetrically available freedom to operate, it means open system, which is essentially commons versus non-commons is the legal institutional parallel of open versus closed. That is to say, you build a system where everybody has symmetrically applied abilities and disabilities around a set of resources, as opposed to one that allows one person to say, you may or may not innovate on this particular set of resources. More generally, we talk about it in terms of open versus closed, whether it's institutional technical organization, and again, critically, at all layers and instantiations of platforms, resources, and pathways. So that means physical, both transport and devices. That means logical, that means platforms, that means content. And here, I'll just very quickly emphasize a few points and then spend the last five minutes highlighting a couple of issues on the digital agenda. So again, what we have when we look at transport wired, and this is back to the point of open internet that David was raising earlier. You can have a competitive structure versus a non-competitive structure. The unbundling and open access rules in Europe push towards a more competitive home broadband market, which means there was less pressure on the behavioral regulation. The reason you're getting net neutrality debates in the US and Canada is because there's not as much of a competitive structure as you're trying to impose a behavioral regulation. In each case, you're getting market discipline as opposed to regulatory discipline to achieve the same result. A platform over one particular bottleneck, connectivity to the home, that does not become a point where you have to ask permission whether or not to innovate, which means you don't have to get into a transaction, which means you can try and fail and fail and then succeed. And there's the question of investment, ownership, private versus municipal. Now we're seeing as the move to ubiquitous mobile connectivity emerges, we're seeing the same thing happening.