In Defense of Qwiki – The Machine That Reads to You

Video artist Ze Frank used to say that it would take him all day long to produce each of his five to ten minute long pieces. That’s not unusual: creation of multi-media content is incredibly time and resource expensive, especially compared to the creation of text content. That’s why I have a lot of interest in today’s public launch of Qwiki – a service that combines speech-to-text and assembled multi-media to create little slideshows based on Wikipedia entries.

Geeks are engaged in heated debate, some arguing that the technology is lightweight, that the product is limited and that the funding of the company to the tune of $8 million by Facebook’s exiled co-founder Eduardo Saverin and others is a sign that Silicon Valley has lost its mind like it did in the original days of the dot com bubble. You know what, though? Mainstream audiences are really excited about Qwiki. I am too.

Sponsor

CNN and PC World gave Qwiki rave reviews today. PC World’s Sara Yin called it “Flipboard meets Wikipedia.” (Links to those sites not included, of course, because mainstream media knows that links to other sites can make you catch cooties.) CNN’s Dan Simon went to the company’s headquarters for an interview and called it a “site that could compete with Google.” (Well, duh, what general search engine doesn’t? Translate that attempt to speak to mainstream audiences though and you’ll see that Simon means Qwiki could be a viable search competitor, presumably because of its wow factor and smarts.) ABC News went with the dorky headline “Get Ready for a Qwiki” and said “this interactive, talking search engine may be the next big Internet sensation.”

Forty thousand people “Liked” Qwiki on Facebook before it even launched, and not just based on the visibility it got from winning a TechCrunch award. People like it a lot.

Yes, Qwiki is simple: it reads the first few sentences of a Wikipedia entry and then slaps up associated images with a Ken Burns effect. A few standard forms of data, like city populations, get turned into slick looking graphs.

Simplicity can be a virtue, though. Watching these awkward little video vignettes may be slower than reading the text, but it’s also more entertaining. I’ve watched probably 50 “Qwikis” now and the quality of the delivery is just good enough to keep me clicking for more. For the bread and circus crowd, Wikipedia read-aloud in a voice just short of the uncanny valley, with images dancing on the screen, could be the difference between Wikipedia content consumed and ignored.

Qwiki’s biggest problem as far as I can tell? That the pieces are too short. I’d love to hear the whole Wikipedia entries read out loud by robots. I’d like to create a playlist of Wikipedia entries for it to read and just let it fly while I’m walking my dogs or riding the city bus.

The web is filled with high-quality written content. Wikipedia, a collection of documents that scores of people have edited on top of each other for years, is a great place for Qwiki to start. That’s not where the company will stop, though. It’s building a service.

Qwiki hopes its technology will be used by all kinds of publishers. “Whether you’re planning a vacation on the web, evaluating restaurants on your phone, or helping with homework in front of the family Google TV, Qwiki is working to deliver information in a format that’s quintessentially human – via storytelling instead of search.” The company lists examples that include real estate listings (good idea) through dating sites (bad idea, sounds like the Tim Ferris dating-for-creeps method, like outsourcing human introductions under false pretense).

The point is, Qwiki wants to let a robot make beautiful movies for you to passively learn about any topic, any text, that you choose. The web is an interactive place, but sometimes it’s good to sit back and enjoy the fruits of that interactivity in a way that asks less of you as a user.

Traditional multi-media content is too expensive to scale to serve the long-tail of would-be consumers. The days of broadcast, mass-media as “the only game in town” are gone. If we’re all going to get multi-media satisfaction, for all our obscure interests, a lot of it is going to need to be created by robots. Not all of it, but a lot of it. There’s no shame in liking that.

Qwiki says it will release an iPad app soon and that could be a big win. Do you want a robot to read well to you, and serve up fancy looking slideshows while it does? I do. I’m not alone, either. Far from it.

Discuss