[wmeredith]

This is what I see coming as the mega corps eat our society alongside software. Look at the regulatory capture the telcos have managed with their negotiated service area monopolies or duopolies. It’s easy to say “buy something else” until there’s nothing else to buy and the product is a requirement for functioning in society (insurance, internet connectivity, healthcare). We need trust busting candidates now.

[jrockway]

I don't think trust busting can make the telco problem go away. You can punish someone for their past wrongdoings, but that doesn't necessarily incentivize anyone to become an ISP after the fact. You have to look at how we invest in infrastructure instead. In most of the country, ISPs are responsible for building their own infrastructure; they have to either dig up streets or beg the local power company for pole access. If they didn't have to do that, many more companies would be interested in being ISPs. As it stands right now, it's too expensive and too risky. Sure, being an ISP is 95% profit once you've paid for the infrastructure, but the payoff period is many decades. Nobody has that risk appetite anymore.

In Manhattan, under every street is a subway where you can put power cables, fiber optic cables, etc. The result is that there are a multitude of business ISPs around, because someone already dug a hole in the ground where their cables can go. The actual stringing of cables is a straightforward procedure; sure you need to buy trucks and cables, but that's a lot cheaper than saying "hey we're closing 5th Ave. for 3 months" and digging the whole thing up. The street was dug up once (in 1891!), and now nobody has to do it again. (OK, the tunnels are pretty full, but it's better than nothing.)

If every street were built like this, it is likely that there would be many more ISPs. The big risks are gone. (There is a secondary problem in that many people still like linear TV, and many TV networks are owned by companies that are also ISPs. AT&T owns HBO. Comcast owns NBC. So that would be the sticking point after there is infrastructure, and sure... have the government break them up. They'll probably be dead long before we have the infrastructure, though.)

The short term solution is probably to invest in better trenching techniques. I've heard of microtrenches, nanotrenches, and picotrenches. Maybe we need femototrenches where you just run a fiber to someone's house through their lawn and cover it up with tape where it crosses the sidewalk. It's better than nothing. But probably still too risky for anyone to attempt.