Microsoft flirted with acquiring Overture, but Bill Gates ultimately nixed the deal. Instead, Yahoo CEO Terry Semel bought the company, launching it into the paid-search business in competition with Google, which had been contesting Overture's patent. The companies ultimately settled and Google agreed to license the patent, but the terms of the settlement were murky. Tellingly, Google's SEC filings mentioned that the license was "fully paid" and "perpetual", but not, as is usually done in such licenses, irrevocable.

Could Yahoo — or Microsoft, if it succeeds in acquiring Yahoo or Yahoo's search business — revoke Google's license, and thereby put billions of dollars at risk? Unless the companies reveal the terms of the agreement, we won't know. It doesn't seem plausible that Google executives would risk massive fraud lawsuits by hiding such a big vulnerability.

But it certainly explains why Microsoft keeps making a run at Yahoo — and why Yahoo insists it's undervalued. But would Steve Ballmer really offer $45 billion of Microsoft shareholders' money to destroy a rival? That's the one part of this outré theory that really fits.

(Photo by emigh)

Remember when Bill Gross launched the search engine GoTo.com, later Overture, and shocked the industry by ranking listings by how much advertisers were willing to pay? At the time, his pay-per-click model was ripped apart as cynical and sleazy. But paid search ads, it turns out, were the right idea. The willingness of an advertiser to pay, among other factors, was actually a useful criterion for determining whether an ad might be effective, and it's a practice that's making billions of dollars for Google.

I think there's a similar effect here. There was no downside for TechCrunch40's presenting companies; if they gave a lousy presentation, all they wasted was the audience's time. One of the best-received startups was MusicShake. Note that MusicShake is based in South Korea, which means that its founders flew across the Pacific to make their debut, at considerable expense. They may not have paid Calacanis and Arrington anything, but unlike most of their peers, they actually had skin in the game.

Contrast that to the sloppy demos given by some Bay Area startuppers who only had to roll out of bed to get there — and looked like they just had.

By contrast, every company at Demo had some money at stake. It might not have been a big risk, but it was enough to make them take the event seriously. Shipley told me earlier this week:

Demo is involved in helping companies launch their products, and there are costs associated with that. That doesn't remove the hurdle of qualifying and being selected to participate. The fee is a commitment for a small company. It does signal a seriousness of intent — it is a bit of a bar for a company, and if you're able to clear that hurdle, that means that there's more at stake and the company is perhaps more viable.

The irony is that when I reported that DemoPit, TechCrunch40's for-pay sideshow that required startups to buy a ticket to the event, Calacanis blusteringly defended his event along much the same lines as Shipley, citing the costs of putting it on. Of course. But by making startups share those costs, Shipley is putting the invisible hand of capitalism to work as a useful sorting function. You know her companies won't waste your time. Why? They have money on the line.