Jason Calacanis argues that the dotCom bust didn't change the overall upward trend in online advertising - and argues further that the uptick is going to keep going:
The real story of Web 2.0 has little to do with the bells and whistles and everything to do with the stunning growth of online advertising. If you look there is a valley between the dotcom spending days (99/2000) and today, but the trend line would be fairly straight if you held a ruler over the 97 to 2006 points--which I do here with the black line. That dotcom overspend, and the dip after it, shouldn't have happened. Those swings were due to the emotional roller coaster of the dotcom bubble on the way up, and four huge events after: the dotcom bubble bursting, the accounting scandals, 9/11, and the brief recession caused by those first three.
He's got charts on his site to back up his idea (along with some caveats about external events impacting it). I think he's mostly right, although I'd add something: we are also moving from mass marketing and broad brush advertising to niche markets and niche advertising. Which means that it will be easy to argue that things are slowing down, even as they increase (but spread out).