With relaunch, Jelly makes another run at Q&A-style “social search”
It uses AI and a routing algorithm to match questions to potential responders.
With a pledge to bring “humanity” back to search, Jelly has relaunched. The Q&A app, or social-search engine if you prefer, originally launched in 2014 and met with limited adoption. Following a refresh and work under the hood, the site/app has returned and is being reintroduced.
The simple idea that resides at the heart of Jelly, as well as many similar efforts that came before, is the notion that people can do a better job of answering questions than a search engine. Despite the logic and coherence of this thesis, nobody has yet pulled it off. That includes a long list of companies and startups, including: Yahoo Answers, Answers.com, Askville (Amazon), ChaCha, Facebook Questions, Keen, Rewarder, Quora and Vark (bought by Google).
Jelly founder (and Twitter co-founder) Biz Stone sees Jelly as the marriage of technology and human altruism. He also believes that Jelly can be more efficient than search:
The average person likely has several if not a dozen or more ordinary questions every day. Think about the time you spend “searching” each day. Those minutes easily turn into hours. Jelly gives you your time back. Enter your question, then return to your life. In its early phase, it may take minutes for Jelly to get you an answer, but Jelly is doing the work.
Like Vark before it, Jelly is using a “routing algorithm” to determine who receives questions. This is intended to match queries with Jelly members best suited to answering them. When you join (not required), you identify areas of expertise and interest. In addition, I also imagine location, historical response times and a number of other factors are considered.
It’s somewhat surprising to me that, despite years of trying, nobody has really developed one of these “help engines” or “question engines” that can go toe to toe with search. But one of the fundamental challenges is scale. You’ve got to have lots of users and responders on the system to create a great experience and reward behavior.
AI and chat bots might help address this basic challenge with curated content and archived answers, where they exist. And Stone says that Jelly is using some AI, but the notion of chat bots is in conflict with the idea of making search more “human.”