A screenshot of Trap.it, a dynamic Web site that delivers personalized content from the Web.
A screenshot of Trap.it, a dynamic Web site that delivers personalized content from the Web. Courtesy/trap.it

Trapit, which shares its origins with Apple's AI personal assistant for the iPhone 4S called Siri, officially launched its public beta Tuesday. The Palo Alto, Calif.-based company, which describes itself as a Pandora for content, offers a discovery system for personalized content based on reading habits, keywords and URLs.

You need technology to discover for you, said Gary Griffiths, Trapit's CEO and co-founder.

Like Siri, Trapit was a by-product of the five-year CALO project, or Cognitive Assistant that Learns and Organizes, which was the largest artificial intelligence program in U.S. history. The $200 million project was co-produced by SRI Ventures and the Department of Defense's innovation arm, the Defense Advanced Research Projects Agency, or DARPA.

Founded in January 2010, Trapit works by providing the user with Trending Traps, which shows popular articles and topics in real-time, and also allows users to create their own traps, which are modules that discover personalized content based on keywords and URLs. For instance, if a user searches for Siri, Trapit will return a page filled with content related to Siri, which may include articles about the iPhone 4S, Apple's mobile division or other voice-activated AI units.

Trapit aims to present users with content outside the traditional news outlets and websites, and allow users to discover stories, reviews and videos from academic blogs or lesser-read websites.

The Web is an echo chamber, the company said on its website. You see the same articles and videos over and over again. Your individual preferences are overlooked by sites and services that take a one-size-fits-all approach, and they privilege the same handful of over-hyped articles and videos above all else. It's much too hard to find the good stuff.

Like Zite or Flipboard, Trapit's algorithm gets smarter over time based on what articles the user clicks on, or whether the user votes their traps up or down. When a user visits an article through Trapit, a banner appears at the top that shows the news source, as well as easy sharing options for Facebook and Twitter.

Since Trapit launched its private beta in June, the company has been tweaking its discovery engine to better help people find relevant articles or stories they are likely to find interesting. The biggest change Trapit has made to its site is the addition of a search bar, which returns search results called traps filled with articles and content related to that topic.

With intelligence of a personalized magazine and the resourcefulness of search, Trapit has positioned itself to directly challenge Google.

We don't think of ourselves as a search engine, says Hank Nothhaft, Jr., the other co-founder of Trapit and the company's chief product officer. We do think the Web is shifting from search to discovery.

Nothhaft and Griffiths equate Google to the yellow pages, in that the search engine is best suited for finding exact results. The Trapit co-founders believe that Google fails, however, as a way to discover new content that you may not have known you wanted.

We're going to be on a collision course with Google, Griffiths said.

Do you think Trapit has a chance of competing with Google in search? Try out Trapit and tell us your impressions in the comments section below.