tag:blogger.com,1999:blog-6569681.post8948408925168026320..comments2024-01-15T13:17:33.771-08:00Comments on Geeking with Greg: Going to CI FooGreg Lindenhttp://www.blogger.com/profile/09216403000599463072noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-6569681.post-91881004507701161092008-02-21T08:51:00.000-08:002008-02-21T08:51:00.000-08:00Hi, Jeremy. Right-o, I think that is the idea beh...Hi, Jeremy. Right-o, I think that is the idea behind the conference, exploring new problems that might be able to be attacked in new ways with these techniques.<BR/><BR/>The particular examples you mention -- reputation and fraud prevention in micro-lending, keeping people informed and engaged in politics -- are great. I am sure there will be many others, probably some completely crazy things, some more practical. I will try to post here about at least some of what is talked about after the event.Greg Lindenhttps://www.blogger.com/profile/09216403000599463072noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-78687596998253747962008-02-21T08:42:00.000-08:002008-02-21T08:42:00.000-08:00what I think Tim is trying to get at is ideas for ...<I>what I think Tim is trying to get at is ideas for new tools that help people help people at very large scale. The goal is to discover the wisdom of the crowd in the noise, "harnessing collective intelligence"</I><BR/><BR/>Yes, yes, I very much understand the wisdom of crowd stuff. I was asking more about the problems, rather than the solutions. The "what" you'd be solving, rather than the "how" you'd be solving it.<BR/><BR/>So for that "what", you mention item recommendation (books, movies, music), web page finding (via delicious, web search, etc.) and collective encyclopaedic authoring.. basically the same set of problems that currently exist. <BR/><BR/>What I was wondering is whether there was going to be any sort of "branching out" in the nature of the problems themselves. At the conference, would there be any thought of how you could apply collective intelligence to the problems that, say, Kiva.org are trying to solve? Ie can you figure out how to use collective intelligence to self-regulate a micro-lending environment? Or to the problem of grassroots political organization, using collective intelligence to help people figure out how and where to make the biggest political different or impact in their communities, nationally, abroad, etc. <BR/><BR/>That's the sort of thing I was looking for.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6569681.post-9524810451574594702008-02-21T07:50:00.000-08:002008-02-21T07:50:00.000-08:00Robert, I wouldn't think of the invites that way. ...Robert, I wouldn't think of the invites that way. Clearly, there are many smarter people and hipper people than me who will not be attending. I think Tim is just trying to invite a diverse group of people with different experiences.Greg Lindenhttps://www.blogger.com/profile/09216403000599463072noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-29317795284327892792008-02-20T22:51:00.000-08:002008-02-20T22:51:00.000-08:00Cool, glad you got to go. If only I was smart and...Cool, glad you got to go. If only I was smart and hip enough to get invited.Robert Vadnaishttps://www.blogger.com/profile/12159736951880674539noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-44450472906845217572008-02-20T18:41:00.000-08:002008-02-20T18:41:00.000-08:00Hi, Jeremy. Frankly, the term "collective intelli...Hi, Jeremy. Frankly, the term "collective intelligence", like AI or machine learning, is incredibly broad. What is and is not collective intelligence is probably just a matter of degree.<BR/><BR/>For example, in collaborative filtering, people rate items, but then machines figure out which other people are similar to this person from those ratings and what information to share between them.<BR/><BR/>In del.icio.us and Flickr, people tag bookmarks or pictures primarily for their own use, but those tags end up having value for others when shared.<BR/><BR/>Wikipedia is a collective effort of many thousands where machines provide tools to ease that effort and search the information.<BR/><BR/>But, hey, even core web search exposes relevant information from humans by eliciting relevance information from the links, clicks, content, and other information people have provided. And, personally, I think web search is collective intelligence. So, where does this stop?<BR/><BR/>I am guessing a bit here, but what I think Tim is trying to get at is ideas for new tools that help people help people at very large scale. The goal is to discover the wisdom of the crowd in the noise, "harnessing collective intelligence", as Tim puts it.Greg Lindenhttps://www.blogger.com/profile/09216403000599463072noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-20386141045283887742008-02-20T18:11:00.000-08:002008-02-20T18:11:00.000-08:00Quick dumb question: Before you build systems that...Quick dumb question: Before you <I>build systems that use "networked computers and humans working together to solve interesting problems."</I>, do you have a sense or idea of the types of interesting problems for which you would be building these systems? <BR/><BR/>Are you thinking about things like web scale image labeling? Or are you thinking even bigger problems?Anonymousnoreply@blogger.com