The discussion so far seems to be largely equivocating on its fair-use understanding of search engine. Torrent trackers take the search methodology a few steps further than Google. For example, Google doesn't mess with IPs in their searches. They resolve through DNS just like everyone else.

Ultimately, though, the real distinction is that Google still relies on a torrent tracker to act as the first-tier search engine for peers.

What I think we really need to do is to start examining ways of completely decentralizing our data transfers. Can anyone imagine a web without servers, just peers?

It's the same thing the computers we are all using do---create useful abstractions and metaphors. Of course states have opinions, they're just the aggregate decisions of the agents acting within rules. Of course information wants to be free---the aggregate human decision is that it's too hard to conceal fully. So why should we use these long sentences every time we want to make a basic point?

Your analogy fails miserably here, because shooting someone is an affront against their _person_, not an abstraction which is the result of their work. Data is data is data is data. Art is more data. Books are more data. We like it more, but they're just data. Data copies freely. The real problem we need to address is how to compensate original authors for the work (i.e. the scarce resource), without violating basic laws of digital media (i.e. a copy has no intrinsic value).