What do you people think, are those 'privacy advocates' who claim that 'even machine-reading of one's mail creates a dangerous precedent' bought (by Google's competition or spam 'providers') or simply morons? Some privacy concerns are legitimate (what Google plans to do with long-term archived mail, what are guarantees that any mining or profiling beyond ad selection will be done only with user's consent etc), but have those guys ever heard of spam filters operated by ISPs? Does that not constitute 'precedent' of 'machine-reading user's mail'?! They should better concentrate on FBI efforts to make FCC force ISPs and backbone operators to re-engineer the whole Internet infrastructure as to make eavesdropping simple, and to pass any costs of that to users.

BTW, I think that Google's plan is to attract users not merely with enormous archival storage and ultra-efficient search eliminating the need for folders and other crude classification, but with nearly-perfect spam filtering, which comes as handy by-product of their superior searching technology.

Concerning those targeted ads (on search pages - I suppose that Gmail ones will be similar), I find them obvious (no fear of confusing them with direct search results), nonintrusive and quite often useful, just as advertised.