On a hunch

Don’t know if you’ve heard about it yet, but hunch.com is open for use by invitation. (I’ve got 3 left as of this post, so email me if you’re interested.)

What is hunch? I could quote something from the site, but let me try and explain it first in my own words: hunch collects data on users of the site, by asking them a number of questions. Then, once a user has a question, she can “ask hunch” about it, and hunch will try and figure out what other people in the same situation have (or would have) chosen. Users can try and train hunch by accepting or rejecting its suggested solutions; in this way hunch grows over time to favour some solutions over others based on the user asking the question.

Here is what the site itself says:

“In 10 questions or less, Hunch will offer you a great solution to your problem, concern or dilemma, on hundreds of topics. Hunch’s answers are based on the collective knowledge of the entire Hunch community, narrowed down to people like you, or just enough like you that you might be mistaken for each other in a dark room. Hunch is designed so that every time it’s used, it learns something new. That means Hunch’s hunches are always getting better.”

That’s all groovy. The catch is the “collective knowledge of the entire Hunch community”. This “collective knowledge” is critical – hunch needs its users to supply it with the content/data it needs. And it is a statistical thing: it needs many users in order to filter out the outlier decisions.

So can it be any good? Are all the solutions it is ever going to suggest for your problem going to be the ho-hum middle of the road ones? Maybe. But then if you think about it, there are not very many unique problems out there. Sure, everyone thinks their specific problem is unique, and surely has never happened to anyone before, and yes that is true, but only to a certain extent. IMHO, what most people need is not the right answer to their problem, but the right questions to ask about the problem.

For example, I just asked hunch whether buying a Kindle is going to be good for me. It asked me various questions, like “do you regularly have more than one book on your nightstand?” (yes), “does your commuteallow you to read?” (yes) and “do you subscribe to major newspapers?”(no). By correlating these answers with all the “Teach hunch about you” questions that I previously answered, it generated a 90% percent match to the solution “yes, a kindle is right for you”.

Once we know which questions to ask, we can weigh them up oursevles. At the least, having hunch throw some questions at you, and seeing whether or not you agree with its suggested solution, can already help you make up your mind. In this case, yes I think I could use a kindle, but I am still shy of the second generation hardware of the kindle2, and would rather wait for the 3rd generation before getting into this technology. So now I can update hunch with my opinion, add pros and cons, so that the next person who is similar to me will get the “no, wait a while” answer instead.

Overall, I like it. It is an emergent design, and therefore interesting!

peace out.