I’ve been reading lots of different things, but a couple of books illustrate the opposite ends of the spectrum of thought about the Internet and its possibilities.
Clifford Stoll’s 1995 Silicon Snake Oil was one of the first manifestations of neo-Luddism. An interesting example of a failed reductio ad absurdam is a critique of estimates of the growth rate of the Internet, where Stoll says

Just by counting network nodes, the Internet is now doubling in size every year. This growth rate can’t continue for long – at this rate, everyone on Earth will be connected by 2003. Impossible

Here we are in 2003, and while it’s not literally true that everyone on Earth is connected, it is true that almost everyone in the developed world has access to the Internet.
Stoll is similarly dismissive of prospects for a revival of public debate on the Internet, taking as his model the dying days of UseNet. He misses almost completely the impact of the Web. Of course, he does not anticipate blogging which, I think, has gone a long way towards fulfilling the early promise of the Internet. Imperfect as blogs are, they contain a lot of well-written and well-argued discussion of a wide range of issues that would have had very little chance of publication in the days before the Internet.

If Stoll was too pessimistic, this was nothing compared to the ludicrous overoptimism of the dotcom boom, chronicled in James Ledbetter’s Starving to Death on $200 Million a Year: The Short Absurd Life of the Industry Standard We have yet to see a really good book on this topic. (Thomas Frank’s One Market Under God, which I reviewed here is easily the best book on the 1990s boom as a whole, but is only peripherally concerned with the dotcoms.