Web discovery engine Trapit launches iPad app — but it’s not another Flipboard

Trapit, a discovery engine for web content from the group behind Siri, is releasing its iPad (s AAPL) app Thursday after launching a public beta web version last November. With the introduction of its tablet app, Trapit plans to compete against news reading apps like Flipboard and Pulse — and thinks it will beat those companies by using artificial intelligence to offer adaptive personalized content. It is also planning to enter into publisher partnerships.

Trapit’s web version has over 25,000 users a day and the company says they spend an average of 16 minutes per visit. The company is based in Palo Alto, Calif. and is venture-backed by Horizons Ventures and SRI International.

As in the web-based version, users start by searching for a subject or URL. They save it as a “trap” and improve its recommendations over time by tapping a thumbs-up or thumbs-down and selecting the reason they don’t like a piece of content. Trapit doesn’t rely on content from social networks, pre-set topics or specific feeds the way apps like Flipboard do.

“Flipboard is essentially an RSS feed,” Trapit co-founder and Chief Product Officer Hank Nothhaft told me. “It’s not bringing anything new to the table. It’s not highly personalized or relevant. We’ve taken the opposite approach. We’re all about user-generated interest, being selfish and really reveling in the things that you like.” Trapit scours content from about 120,000 sources, up from 100,000 last year. In the past few months, the company also cut about 10,000 sources based on user feedback.

The Trapit iPad app lets users share their finds through Twitter, Facebook (s FB), Pinterest and email, and they can use a “read later” button to create a reading list. Other unique iPad features include Retina display optimization, voice recognition and a reading format that “shares the love with the original publisher” of a piece of content.

Other apps “lift, borrow or steal a good chunk of content and text and present it in their interface,” Nothhaft said. “We set out to create a seamless browsing experience” that serves pages within the app but shows their original design, videos, comments and ads and counts as a site visit.

Coming soon: Publishing partnerships and “source-based traps”

Nothhaft said Trapit is making “a lot of inroads with publishers,” just signed its first deal with a magazine and will add premium content soon — not just “the usual subjects like newspapers with paywalls,” but longer-tail content centered around topics like crafts, cooking and DIY. While a revenue model isn’t decided yet, users could upgrade to the premium content or just buy it buy the piece.

That is the opposite approach that magazine joint venture Next Issue Media takes with its “all-you-can-read” tablet magazine subscriptions and focus on popular publications. “Big magazine content doesn’t make sense in a Trapit context,” Nothhaft said, describing approaches like Next Issue’s as “antiquated” on a tablet platform. “We like the idea of unbundling and breaking a [magazine] issue apart.”

In addition, Trapit plans to add the ability to pull in all posts from a given website. (When I wrote about the company last November, I mentioned that Trapit could not yet replace my Google Reader(s GOOG) because I want to read all posts on a few sites and don’t trust an algorithm not to miss something.) “This makes sense to us even though it’s not the core of what we do,” Nothhaft said. “The combination of deep personalization with the ability to follow the sources you really love eliminates the need for multiple [content consumption] apps.”