I played with over-hyped new search engine Cuil.com off and on throughout the day, and I’m amazed at how poorly and inconsistently it worked. There appear to be serious architectural flaws that resulted in inconsistent and unpredictable results.

John Dvorak has a scathing review slamming Cruil for poor search results.

And Juan Carlos Perez points out that Cuil’s poor first impression could do lasting damage to the brand the owners are trying to create.

One interesting thing I noted was that as the day wore on, Wikipedia articles started suddenly appearing high up in the search results where they had been conspicuously absent before (something a number of other bloggers pointed out). That’s a positive sign — when you’re researching a famous person, you expect a competent search engine to return an encyclopedia entry fairly high up in the results.

But even if Cuil gets its act together, it doesn’t seem to bring anything compelling to the search engine space. Maybe I’ll change my mind once the stability and load balancing issues are sorted out.