I'm going to preface this blog post by noting that it's somewhat more technical than political, so if that is uninteresting, feel free to ignore.
I've been thinking a bit recently about what the internet needs moving forward. Specifically, I see a few looming problems which threaten to harm, if not effectively destroy, the internet as we currently know it. First, there is the problem of pervasive government monitoring, which threatens free communication of ideas. Second, there is the attack on net neutrality, which threatens to make only the content which the telcom providers sell viable on the internet. Both of these are serious problems, and although it's conceptually possible that they would have legislative solutions, that presumes the laughable premise that the government works on behalf of the people (as opposed to itself and/or large corporations who bribe them), which is obviously false. Therefore, it's prudent to look for a technological solution for these problems, if possible.
Fortunately, I do think a technological solution is possible (albeit at the expense of performance). Specifically, here's what I think we (the people) need: open source router firmware (dd-wrt style) which implements automatic multi-channel TOR communication, and passes all traffic through them, while itself serving as a TOR relay node (and optional exit node).
How would this work? TOR, by its nature, obfuscates traffic destinations, and encrypts traffic. It is, currently, one of the best defenses against network surveillance, both by individuals and governments. By passing data through multiple channels, you could guard against traffic analysis [at a node beyond the ISP level], and having each router act as a relay node would ensure wide distribution of nodes.
Moreover, this would effectively make net neutrality a de facto standard. With all network traffic sent through TOR relays, the ISP's would have no method of discerning the protocol or destination of traffic, and thus no metric upon which to bias the bandwidth. It's possible that they could apply a QoS to throttle TOR traffic itself (as some currently do with torrent traffic, for example), but with everything going over TOR, the ISP's would need to adjust their strategies or lose customers (and/or be subject to false-advertising claims with respect to their bandwidth). If enough people were using the firmware, it would become the de facto standard, and ISP's would be forced to live with it.
What about cost? Well, that's not insignificant: it would roughly triple the amount of network traffic necessary for the same communications, not counting encryption processing. It's even worse than that, though: multi-casting would become effectively impossible, and some elements of malicious usage detection and traffic would be rendered effectively impossible. Those would be the necessary costs of implementing a technological solution to a problem which would be much more efficiently solved at a political level, if not for the irreparably corrupt government we must deal with. Still, I think it would be worth the cost going forward, if the alternative is the internet as we know it ceasing to exist.
As always, I'd be interested in other opinions, either on this or other potential solutions. I don't know if something like this will happen, but given the various alternatives, I really hope that creative, smart people come up with some solution, before it's too late.