Couldn't agree more. Had to copy it here cause of the mention of AJAX and KoZmo. From this TechDirt article:

There's been lots of talk about the possibility of "Bubble 2.0," which (so far) appears to be taking a lot of ideas from the original dot com bubble, and using dynamic HTML to make it look cooler (it moves! in a browser!). While it's true that some of these second attempts are doing things much more intelligently this time around, an increasing number are not. Once the money starts flowing (and, if you hadn't noticed, it's flowing), it seems that "lessons learned" get tossed out in favor of doing anything possible to jump on the gravy train. The only problem is that we've seen how this train ride ends, and it's a train wreck that we'd all probably be better off avoiding. Still, if you want to revisit the signposts of years past, here are two more: it's suddenly fashionable again to put your startup in "stealth mode." Stealth mode is a silly term for it. Being quiet about what you're doing certainly can make sense -- especially early on when you're still figuring out the details of your messaging and positioning. However, too many startups use stealth mode as a buzz builder -- as if not telling people what you do deserves more attention than having an actual product. It's that second use of stealth mode that tends to cause over-inflated expectations. The second bubble sign is when the failures of the original dot com bubble return from the dead to try to make it again. In this case, it's "boo.com," whose only claim to fame these days was how spectacularly it flamed out. However, in a bubble era when "any publicity is good publicity," the larger the failure you're trying to resurrect, the more publicity you'll likely receive. So, come on Webvan, Kozmo and others. It's time to come on back. All you need to do is use AJAX and it'll make people forget to ask if you have a business model.