Right, so I agree that the engineering challenges aren't very great, but I would argue that little of the current infrastructure was designed with the ability t

Right, so I agree that the engineering challenges aren't very great, but I would argue that little of the current infrastructure was designed with the ability to de-prioritize packets based on their origin, and so there'd be a pretty massive replacement of hardware in order. I'd also argue the above as a more realistic attribution of motivation than that they're worried people will just want less Internet. I don't think the hardware companies are shills, or that they're evil, I just think they're myopically focused on their next product cycle. The point you raise about the size disparity between Netflix and Comcast is precient, but I see it from the other direction - what would Comcast and their partners start doing to market entrants with this kind of strategy? If I'm Google and I've got a comfortable relationship with Comcast, what's to stop me from asking them, as part of a larger deal, to severely degrade or outright block new entrants that might threaten me? We've seen lots of companies do this (Microsoft and the various computer manufacturers (Compaq et al.) in the 90's comes immediately to mind). The trouble with this scenario is that it takes a long time to notice anything is wrong, and so you don't get much market pushback. Then 10 years go by and you realize that no one's challenged the incumbents in a long time, and it's not because they're any good at what they do. > My future Internet connectivity was supposed to be way better than my present. I don't think that's safe to say at all, considering it's been flatlined for so long: http://blog.nprg.com/wp-content/uploads/2012/07/Google-Fiber-Three-components.png http://blog.nprg.com/wp-content/uploads/2012/07/Google-Fiber... Which is why I don't think the light-touch approach is warranted in general anymore. There's clearly big structural problems here. Having said that, I think we'd agree that the above chart is mostly a last-mile issue. The NN thing is more future-oriented, which is riskier, but also hopefully avoids a situation like the last mile issue where things are so entrenched that it's hard to imagine fixing it.