When they first announced it, I was certain Knol was simply a way for Google steal some of the attention away from Wikia search -- and I'm still sure that's what was actually happening. Google doesn't normally pre-announce applications months in advance. After a long wait, Google finally released Knol, and the blogosphere has been very active on the topic.
Most people don't believe this is a direct threat to Wikipedia -- and technically I guess it's not. You can create openly collaborative Knol's (Google defines that as a single unit of knowledge), but more often than not, they are closely moderated by the author since their name (and reputation) is permanently attached to it. Instead of anonymous, and potentially incorrect contributions, Knol forces people to identify themselves (or be marked as "unconfirmed"), and put their reputation on the line. This means higher quality articles that are less prone to error.
If you don't like an article that someone has created, give it a low rank and start your own on the same topic. That's right, Knol doesn't enforce a "one Knol per topic" rule -- another significant difference between the two services.
Even though Knol is not direct competition, it could very well suck the life out of Wikipedia indirectly. No, it won't happen over night, and yes, people who have been contributing to Wikipedia and are passionate about it will continue to do so. Unfortunately for Wikipedia, Google has the means to pay authors for their contributions. I wonder what's stopping someone from ripping off all Wikipedia content and profiting from it?