Venture Beat has an interesting story about a stealth start-up called Powerset that is apparently on the verge of raising some serious venture capital based on the promise of a new Google-killer search engine using artificial intelligence and natural language. Given Google's dominance from a market share, technology and branding perspective, the idea of actually going to head to head with it seems like a very Don Quixote-like exercise. Yet the search engine market continues to capture the imagination of bright people and investors intent on coming up with a better mouse trap. Some new search players such as Collarity are intent on taking on Google directly, while others such as Kayak.com (travel) ZoomInfo (people), Topix (news) LivePlasma.com (music, movies) and RedToronto.com (local search) are focused on picking a fertile niche to establish a foothold. The question in search is: what would it take to knock off Google? Is it possible to develop a search engine that can produce better results than Google and, at the same time, capture the imagination of Internet users? The idea of developing a search engine that offers better results and/or provides a better interface is entirely possible. Imagine a user-friendly search engine that uses graphics rather than text. The hard part will be weaning people off Google or at least offering a new option that encourages Google-lovers to give something new a chance. When Google appeared on the scene, the established players were Lycos and Excite. They were pretty good search engines but not great, which gave Google the opportunity to quickly build a following as more people discovered it. While it would be wrong to dismiss Powerset's prospects, out-Googling Google is an enormous challenge but if it keeps Google (and Yahoo and Microsoft) honest and pushes the technology envelope, bring it on.
Update: Amazon's A9.com search engine has dropped some of its features. Speaking of Google, it has launched a new search engine called Searchmash that appears to be a trial balloon project. In other words, Google could take whatever works well in Searchmash and adopt it. Google Watch has a post on 10 reasons why it's almost impossible to compete with Google.