How Twitter should make money (The Twitter supercomputer)

May 28, 2008

OK let me set this up - I didn’t really “get twitter” when I first started using it. It was just a toy until I started having interesting interactions with my friends and my time spent having “useless banter chats” on IM went down that I got more into it. It started getting cool when I started chatting with some people out side of my regular social circles - now they are regular contacts. So that’s interesting, that’s new but it seems to be a black hole for VC money right now. Twitter will most likely attempt to go the Ad model route to stay afloat. While that will probably earn some coin, I doubt directing small text spam ads to users and banners will really accomplish anything for all parties involved.

So to abstract it - Twitter is a disconnected asynchronous open social communication tool where all the users involved use it for small periods of time and are only really into inputting and consuming small bits of information. I think most users of this system are pretty cutting edge (which means they don’t care about ads) and I also think most of them are willing to trade small fractions of time /effort in order to keep this thing afloat.

OK so is it a stretch to call twitter a neural network?

Information passes amongst nodes (that’s us) from one party to another. Clusters of nodes (social circles) can interact with each other and information can pass to other circles. Why not leverage Twitter as .. well.. a as a super computer. Kinda like SETI@HOME but with a Amazon Mechanical Turk twist. Yea yea yea, I’m bringing up the “this + AMT = awesome” thing again. However after seeing this TED talk - I’m now 100% sure this concept is the future of the internet, in terms of big social impact.

Here is an interaction I just dreamed up if say I told Twitter I was good at both french and english (which I’m not). It could find a fragment of text needed to be translated and send it to me. My response will be seen by all of my friend and has the potential to span across multiple social clusters that could result in better and better responses.

Twitter Agent (direct message to me) Please translate to french: Press F5 to enter debug mode
Me, who only knows a little bit of french from school: @twitcore Le press f5 to entra mode du debug
My french speaking friend that knows French Fluently: @twitcore @bpandrew pressez f5 pour entrer corrigent le mode
My french speaking friends friend who knows French AND software : @twitcore @bpandrew [please imagine a even better translation here, the above one was from babelfish]
Friend of friend of friend who is even better than friend above that agrees with the above translation : @twitcore @bpandrew [i dunno, a yes/no? maybe a re-tweet of the answer above if its perfect?]

Maybe french wasn’t the best example but hopefully you got what went down there. What’s been going on amongst my cluster of friends are questions about the Javascript API jQuery, I’ll post a line of code and friends have been sending me back more optimized versions. Maybe even in addition to improving information friends can could “vote” on answers to provide a sort of social validation. That way by the time the twitter core gets and answer back, it will have multiple revisions plus a host of votes from many users.

Twitter could go out and started scoring deals to get information processed the based on “interview” profiles of their user base - figure out who is good at what, fragment out small chunks of work and pass it along. I don’t know - this probably has lots of holes in it, but that doesn’t mean the concept should be passed over. I just see the ad revenue model to be like putting cheap all-season tires on a F1 race car. It gets the job done but a lot of potential is going to waste.

Tags: , ,

2 Responses

  1. With respect to Twitter making money, they are already doing advertising on their Japanese site:

    http://news.cnet.com/8301-13772_3-9926331-52.html

    You might also find Jason Calacanis’s thoughts about Twitter’s business model options interesting: http://www.calacanis.com/2008/01/02/the-three-business-models-that-make-twitter-a-billion-dollar-bus/

    While I’m equally fascinated by Amazon MT, I think Twitter may be more akin to something like LinkedIn’s Answers feature. Where there is some structure for posing and responding to questions put to your own and other social groups. It seems quite doable that this functionality could be built on top of Twitter as it stands now as a karma/whuffie type system.

    That being said, I am in definite agreement that doing something to capitalize on the unique human and social aspects of Twitter would likely serve them better in the long term than slapping some advertising across the site would.

  2. @Michael

    Jasons ideas are bang on about critical mass - revenue doesn’t really matter for them right now I guess - having solid uptime is more important in the grand scheme of things. Once they’ve hit critical mass they can pretty much do anything with degrees of success.

    re: The LinkedIn’s answers thing - I stopped using that site a while ago so I haven’t really seen it in action - is it just like yahoos answers? or can questions be pushed out towards users/groups?

    Thanks for responding I knew you’d have some valuable insight on this topic.

Leave a Reply