I'm looking for a partner or two (here are some of my ideas, feel free to add your own here)

I'm interested in ways to make people scale. For example, one might imagine a service that allows people to construct a chatbot for themselves that they could unleash onto the net to spread their views. Yet another possibility is to have such a chatbot constructed in a collaborative way by many people. The interaction between such chatbots is also interesting. For example, you could have an online dating site where you can go on thousands of dates per day by having your chatbot interact with other chatbots. Promising interactions can then be brought to your attention and you may then decide to go on a real date.

This is my current startup attempt: http://forwardingtree.com. Also see: http://groups.google.com/group/forwarding-tree. Note that personalization is just another way to control spam. One might imagine an extension of the social network approach used that would apply to discussion forums for example, even at the level of individual threads/posters. How would you personalize a discussion and yet still retain something that is coherent?

I'll present some other ideas here soon.

I am interested in getting free labor out of users in the style of the espgame.org. For more on this, see: http://video.google.com/videoplay?docid=-8246463980976635143 An obvious step would be to extend the ESP Game to video. Of course, the service need not be a game. For example, consider two groups of people: those who wish to grammar check their documents and those who wish to learn English. To grammar check, one would submit the document and users trying to learn English can try to find errors in it and provide corrections. General agreement about certain sorts of errors might be viewed as making it more likely that those in fact are errors. Moreover, one can also look for agreement among corrections. People learning English would benefit by getting back feedback about how well their errors and corrections correlate with those from others. One could of course do something similar with code design say.

I'm interested in novel uses of prediction markets. For example, one can create an online dating service where people bet on whom they believe would make good couples. People might then take this as a suggestion as to whom to consider for a date. A similar thing can be done with business relationships, academic collaborations, etc.

I'm interested in a service for creating or inventing things in a collaborative way. For example, consider using human-based genetic algorithms to create daily news shows, perhaps in the style of this site: http://www.newsatseven.com (but using people rather than computers). As another example, one might build a service where users invent board games. Again, this could be done using human-based genetic algorithms. Users of the service would be able to modify rules to try to make a game more fun. A fitness function would take into account how popular various game variants are among players. Note that users of such a service need not be programmers: the domain is sufficiently restricted so as to make visual point-and-click rule creation/modification possible.

I think it would be interesting to provide a way to give people advice in real-time via the web. For example, you might watch a sports game on the web and give advice to individual players about what they might do next in real-time. So the spectators replace the coaches here.

I am interested in collective intelligence: http://en.wikipedia.org/wiki/Collective_intelligence. How would you build a service where thousands of people collaborative to solve a problem, create a work of art, or play a game (e.g., chess)?

I'm interested in ways to make more novel startups succeed. The problem here is that the average web surfer has a very short attention span. And so it is hard for anything novel to get much traffic. One possible solution is to try to extend the attention span of the average web surfer much as you would with young children. Namely, present an idea in a cute and funny way. See for example: http://www.hitachigst.com/hdd/research/recording_head/pr/PerpendicularAnimation.html . By extending their attention span in this way, you might be able to get them to understand and appreciate your novel service.

I am interested in seeing what happens when you apply intentional programming to web search:

http://youtube.com/watch?v=tSnnfUj1XCQ

http://youtube.com/watch?v=ZZDwB4-DPXE

The idea is to open up your search engine so that anyone can submit code for domain specific query entry, search, ranking, and presentation of results. Moreover, such code submissions would compete with each other, with more effective ones rising in popularity in their respective domains.

Unlike APIs that are currently available, the submissions made by users here would make up the core feature set of the service.

I am interested in giving people ways to track their influence on others. Forwarding Tree's URL tracking is all about that: http://groups.google.com/group/forwarding-tree/browse_thread/thread/ad998b99b9e3b7a8

URL tracking allows you to see what happens to those URLs that you forward to others, post on your blog, submit to a social news site, etc. You can see how the URL propagates around the net in real-time.

In other contexts, one might show people their influence on wikipedia articles, search engine rankings, open source code, etc.

What about advertising using smart mobs? A web service would allow a company to specify the logistics of the event and how much it is willing to pay. The company could also track the event in real-time and make changes as required. Advertising could be subtle such as in the form of a mob that forms in a shopping center carrying several products from that company. And advertising could be more blatant: a mob might form to create the logo of the company say.

I'm interested in understanding people more deeply and learning from them by watching what they do and how they think. One idea that I tried is a service for broadcasting your computer screen, just as if it were a TV show. See my work on VNCCasts: http://www.mobuzztv.com/uk/shows/vnccast_my_scrap_beckett_sculpture/. (The report is a bit misleading though as they apparently mistook the sample channels I created as ones actually used by others.)

I'm interested in showing people data that they don't normally see. I mentioned VNCCasts in another post. Another example would be to show the web traffic at a web site to everyone in real-time. This might be done using something like http://www.visitorville.com/. Or maybe you would show actual sessions in progress. But whatever you do, the point is to make the site look more active and interesting. It's a way to promote the site.

I am interested in more intelligent ads. Make the ad so intelligent and interactive that people would want to play with it, just to test its boundaries. Building a better ad now reduces to coming up with the better AI. Taking this further, you might allow people to combine two ads to see how they will interact with each other. For example, you might combine an Apple ad with a Microsoft one to see which AI will win out.