Flickr Co-founder Unveils Her New Startup: Hunch
People have been whispering about a new web application in development called Hunch. Today, Flickr co-founder and Hunch head honcho Caterina Fake divulged some more details about the new project on her blog.
The new project aims to become a site that can help anyone make a decision about anything. The way it will do this is through the application of decision trees that are created by contributing users. Using decision trees in expert systems is nothing new, but applying that idea to a crowdsourcing model might possibly be a stroke of genius. Think Aardvark meets Wikipedia and you start to get the idea.
Hunch is still in closed beta, but is accepting requests for invites. We're thinking you will want to sign up, though, after you hear what Caterina says about it:
Look. Decision-making is difficult, and decisions have to be made constantly. What should I be for Halloween? Do I need a Porsche? Does my hipster facial hair make me look stupid? Is Phoenix a good place to retire? Whom should I vote for? What toe ring should I buy?
It's dark and lonely work. Coin-flipping, I Ching consultation, closing your eyes and jumping, postponing the inevitable, Rock-Paper-Scissors, and asking your sister are all time-honored means of coming to a decision -- and yet we think there's room for one more: Hunch.
She adds that a lot of content in Hunch is going to be generated by its user base. Do you know the right questions to ask to help someone pick out the right pair of shoes for hiking, or what cell phone to buy? With Hunch, you will be able to get in on the ground floor and know that your contributions will help many people get the right answer to their question.
While we know very little about the inner workings of Hunch, it apparently combines decision trees with a fair amount of end user personalization in the form of questions it asks people visiting the site. These questions allow Hunch to form affinities with other users who ask similar questions. On the back end, contributors will be able to create topic areas (called Super Questions) and add questions and results underneath those topics. How much control you will have or how the interface looks for this we aren't sure yet.
Caterina also says that there is room for the site to make money, by including Super Question areas that are affiliated with commercial products or services, but that part is not being rushed as they want to get the core functionality working just right.
We think the potential for the idea behind Hunch is huge. If you look at another very famous crowdsourced project, Wikipedia, and combine that with the sheer utility of the application Aardvark (which lets you pose questions to an extensive network of Aardvark-managed instant messaging contacts; our review here), the sky really is the limit for how much this tool can grow in usefulness and popularity.
Self-titled photo courtesy of Caterina Fake on Flickr.
Share
Facebook
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
Sounds like a very cunning plan. Facilitating decvision making is potentially a very sweet spot. Props to Caterina. Great name too.
How does it differ from FriendFeed, Facebook, or Twitter? :|
Posted by: Mona Nomura
|
March 27, 2009 3:54 PM
It’s a site for the indecisive. Now that's a huge market! It's a clever name and idea. I have to admit that I used to be very critical of Flickr, although it has improved tremendously. Flickr is the only good reason I have not to try it. I'm not knocking her success either. Nevertheless, I'm also not confusing it with the ability to create a quality service. It's too bad they won't be able to help me decide if I want to try it or not. I trust your opinion, but you haven't tried it either. I'll have to wait for your review on this one.
This is a dumb idea. Its the virtual 8 ball and just like the 8 ball its a novelty and usually ends up in the trash.
Mona doesn't know how to read, apparently.
uh, if she sold flickr to yahoo for a giant pile of filthy lucre...why does she need to do this?
there is a problem with crowd sourcing, initially Hunch will be tailored for US people, there are other cultures and people in the world, who are different than what Americans like! so basically in the 1st year, this system will be useless for non American users.
@Mike:
"This is a dumb idea"
Strongly disagree. ppl on twitter are forever throwing up tweets like
"Dear lazyweb, should I ..." "... is this ..." and so on.
All it does is formalise that and put it into a struture. It follows the collective intelligence model. The wisdom of crowds, as it were.
It'll turn _very_ interesting once the marketers get hold of it to push their own products (watch for the API's to appear so marketers can have bots alerting them to questions on, say, shoes where they can swing into action to promote their own product). Marketers are already all over Twitter corrupting it with indirect promotion (I block anyone who follows 1000+ ppl) and if they get their hands on this, it'll start raising all sorts of questions about the true worth of Web2.0
bring in on I say, lets see what happens.
No way!
I had a hunch someone would come up with this idea too.
Unfortunately I don't have the resources Caterina has to make it happen. I developed a detailed plan a couple of years ago to build a similar service. I guess coming up with ideas is not the problem but to be able to successfully execute them.
My execution plan differs quite a bit from what I see she is planning to do.
Having the system to become smarter overtime is a very challenging thing to do and only time will tell if hunch can actually deliver useful hunchs...
Also, Hunch has to make sure they are not liable for people's mistakes based on advice obtained from their service.
Imagine that you make a decision based on Hunch that turns out to be horrible or even deadly on the long run.. you can see the law suits coming!
The possibilities are endless though because decision making is a universal need, everyone has to make decisions and with so many options today it becomes more and more difficult to have the time to evaluate all possible choices.. A system that helps you doing the basic criteria crunching could be extremely useful for everyone.
I really hope their execution becomes successful and in the meantime I will try my own attempt.
I'm playing with it now, and I like it, I think it's groundbreaking.
I have some invites too.
I think this has lots of potential. Just taking someone through the questioning phase may be enough. Parts of it remind me of a Windows program called IdeaFisher from ages ago that had a library of topics and questions.
Could be a viable competitor to Yahoo Answers and other help forums.
Posted by: Vezquex (beta)
|
March 27, 2009 6:12 PM
The glowing reviews on VB and RWW make my stomach turn. Mona is right, the only difference is the purpose of the site is very specific, and it doesn't have the friend/follower/subscriber/etc mechanism incorporated (maybe that's on the road map?). There isn't one critical sentence in either of the write-ups. This would be a first in the history of product development: initial product launch is flawless without any shortcomings. These blogs are just happy TC didn't get the scoop? Credibility>shitter.
Posted by: coldbrew
|
March 27, 2009 6:43 PM
@Coldbrew, you've been reading and commenting here for a long time and I think you've called our credibility shot a few times. You keep coming back though and some of your comments are very helpful. Not this one so much. If you'd read Phil's post a little closer you'd see that he says we haven't tested the thing yet. If we had I'm sure we'd have some critical things to say about it, but the even the founder says it is in a super early stage.
As for differentiation, like Phil says - this appears to be based on collaboratively written decision trees. How much more different from Twitter/FF could it get on that level? I still haven't gotten an invite yet so I can't say for sure, of course.
You should go have a cold one and send back the more helpful @coldbrew.
Sounds like a good idea, even if only for its entertainment value.
But shouldn't Caterina be referred to as head "huncho" rather than honcho? :)
O boy now the internet will be making decisions for us. I think its a cool idea but good/practical in the end? Not so sure.
But I can see huge potential from the entertainment stand point on small decisions.
> Hunch is Aardvark meets Wikipedia and you start to get the idea.....Please stop with the vacuous hype. I'm using Aardvark and the crowdsourcing model works OK at best. I expect Hunch to me nothing more than another shiny object with a new marketing wrapper.
*im thinking the same way too..
"How does it differ from FriendFeed, Facebook, or Twitter? :|
Posted by: Mona Posted on FriendFeed | March 27, 2009 3:54 PM"
What I think is "cunning" and "brilliant" about this is how people can get other, unpaid lackeys to do their labor ("a lot of content in Hunch is going to be generated by its user base") -- let us add UNPAID user base -- whilst the person responsible for the general idea sits back and makes a pile of money.
I guess, yes, in today's internet world, that DOES constitute brilliance.
Hunch is interesting and is very helpful. Caterina Fake has a superb concept for establishing this site.
This site is developing decision trees, not making decisions. It is developing the process of making decisions for people and helping them make them intelligently. It is the kind of site that might make your head hurt if you give it too much sway in your decision making.
Let's start with a simple question from Barack: Should I spend trillions of dollars bailing out banks?
There are going to be trees that are very compelling.
Little decisions will be fun.
Big ones may change the world!
I think it could be very comforting to know there is a place to go and ask for advice for something that is really bothering you and get a long answer. It could help. There is alot of talent on the web. Like the name. I would consider using it especially if I find talented people there to help with answers or atleast give me a hunch about it
i am very skeptic about its success..... i hope it does wonders... best of lulck
Sounds very interesting. Reminds me of times I've found articles on ehow.com with steps for making a decision, like "how to choose a string trimmer".
I recently soft launched a little tool I put together to help people use "decision matrices" to make decisions. (I actually built it to help me figure out what kind of string trimmer to buy") I think decision trees and decision matrices would be useful for different types of decisions. 'wonder if Hunch plans to incorporate other kinds of decision-making tools.
My decision matrix tool is hosted at http://www.topwhatever.com
Fail
Credibility is not binary, and yes, writing press releases for companies or products affects how RWW is viewed as an objective source of info. WRT writing w/o testing, I believe this fact makes the case that this post is equivalent to company-released PR. Even commenter "barry.b" above points out the probable monetization strategy and the issues that will arise as a result (I doubt barry.b has "tested the thing").
Posted by: coldbrew
|
March 28, 2009 11:35 AM
Hunch was intentionally developed w/ a very specific purpose for use in making decisions but ppl already use the more flexible FF/Twitter/etc. services for decision help. Scoble asks "what's the best blah?" all the time & generates quite a debate where people share their knowledge and experiences. These discussions help many ppl make decisions on all kinds of things. I understand the concept of developing a tool to "do one thing and do it well"; Hunch just dumbs down a process many of us enjoy.
Posted by: coldbrew
|
March 28, 2009 11:43 AM
I'm skeptical but will be giving a try never-the-less. Decision trees and Prolog bsck-chaining expert Systems both had/have wins but this attempt seems to be an attempt to weave heuristics / common sense into the process ... I don't see how it can provide the requisite variety to accomplish all the aims it claims to have (being based on tree structure at the core) but it might be a great start for someone with an area of interest where they don't even know the right questions to consider...
Posted by: David HC Soul
|
March 28, 2009 11:52 AM
Stroke of genius? Really? http://www.unfuse.com
Unfuse was released early 2008.
Good idea. Difficult to execute and fine tuned. It also made me think of Yahoo! Answer in which the model of simple Q&A is more straight forward than a decision tree type of questions. But the quality of intelligence it collected from the crowd seems not that impressive.
From liability point of view, Yahoo Answer shares the same problem. It seems that its not that serious concern.
Interesting idea, I just read about this web site with a similar concept - www.voteonmylife.com. Where people vote on other people's decisions.
The concept sounds interesting, but it's not new. Check out: www.voteonmylife.com. They have been up for almost a month now.
I don't know if I would use this or not... decision trees are great, but why not just go to Yahoo Answers and get opinions direct and specific from real people with expertise instead of a vague tree.
Sounds like a semantic Yahoo Answers who wants to build an expert system in a collective intelligence counterpart to Wolfram's scientific inference engine. Very nice !
it's a very pretty package. and, it did help me decide what i should eat for lunch. ;) i will continue poking at it to see how else i can benefit.
my experience and thoughts thus far, http://web-poet.com/2009/03/28/hunch/
VERY shallow so far. I'll try again when most contributors are on their third or forth try and/or iteration of tree (or when some natural selection has set up a landscape where the 'good' trees are evident from usage.
Posted by: David HC Soul
|
March 29, 2009 2:02 PM
If the Wolfram inference engine can do what it claims it can in the next month or so, then Hunch is toast. It can't compete with a superior engine such as the one that Stephen developed.
I think it's great to see all this innovation around Q&A. It just goes to show that one of the main things people want to do online is simply ask questions and get them answered.
The concept behind Hunch is good but obviously challenging and in need of testing to see whether it works and whether people like the experience. I'll be trying it.
The thing I find strange is the fact that you can't simply ask questions of the people (employees, visitors, customers) at any web site you visit. We tried to solve that problem at YouSaidIt (www.yousaidit.com) by letting people add Q&A to their site.
I really don't see this as 'innovation'. I suppose adding in an affinity component is clever, but really, it just turns a massive set of q&a's into a slightly more limited q&a's that might be more suitable for you - which if you think about - is a bit like reversing the 'crowd-source' effect. It sounds like they're just building a slightly more complicated quiz system ("Who should I vote for? Really?").
I remember hitting up a website in the 90s that guessed what animal you were thinking of based on decision trees, and if it couldn't figure it out, you were expected to add the questions that would have solved it for you. Yeah, it worked, but it didn't feel particularly clever.
And, as has been said previously, if WolframAlpha is what they claim it to be, Hunch hasn't got a snowballs chance.
Not really an original idea, though. The site http://www.decide-now.com has been running a similar spoof about decision making for many years.
Phil -
Thank you for highlighting this new venture. This is an interesting idea and for people who may not be great critical thinkers as this would aide them in working through a decision making process. It will be interesting to see if this concept gains traction and if it can be monetized.
I will be watching.
James
http://Twitter.com/AskJamesHolmes
Sounds a bit like answer.com but it might be better depending on how they implement it.
This sounds like a fantastic idea.
Best,
Tim Rad
http://twitter.com/supercamp
Caterina Fake is anything but FAKE. :-)
Finally a genuine attempt at trying to help people with their lives - instead of filling zillions of megabytes with useless twittering and facebooking of "Mam, I just made a big poop, ain't I great?"
Fact is we all spend an aweful lot of time in front of our one-eyed medusa machines. We all will be better off if we start using that time for REAL life.
Brava, Caterina!
Thanks for the beta invite. I just kicked the wheels a bit and thought I would come back here to see what other RWWs thought.
I don't have experience of many of the web apps mentioned above so cannot comment on which is best, what I can say is that the biggest difference between this and Twitter/FF/FB IMHO (going back to the original question by @Mona) is that the latter will only help you answer questions if you have reached some sort of tipping point in terms of followers. We know that the majority of users will not get to such a tipping point (the average RWW user is an exception to this statement I imagine.)
It is a platform and will depend on UGC to make it valuable, just like Flickr. If you don't like that concept then stay away. However, if you like to contribute to something that is bigger than you form time to time (i.e. like commenting on a blog like this) then you may enjoy using it.
On balance I think I will probably test drive it a bit more and then see whether it sticks.
Flickr is the only good reason I have not to try it. I'm not knocking her success either. Nevertheless, I'm also not confusing it with the ability to create a quality service. It's too bad they won't be able to help me decide if I want to try it or not. I trust your opinion, but you haven't tried it either. I'll have to wait for your review on this one.
I think this is the best lines i have found in your article.
"The new project aims to become a site that can help anyone make a decision about anything. The way it will do this is through the application of decision trees that are created by contributing users. Using decision trees in expert systems is nothing new, but applying that idea to a crowdsourcing model might possibly be a stroke of genius."