Hola folks,

I figured I'd ask if anybody has any input on 'Citybot' - a small project
I'm trying to do.  I may present it at wherecamp here in PDX in a few hours
although the project itself is only a few weeks old and hasn't had more than
4 or 5 hours of code put into it.

The background on Citybot is that it is basically is a minimalist attempt to
build a social network on top of twitter that lets people barter and trade
things more easily.  I have it hosted at http://citybot.org and I have few
posts and comments on it at http://makerlab.org as well as some help / about
information on the site itself.

The big picture is that I'm interested in "signaling networks" between
people.  I want to let people see through walls.  My interest in social
cartography stems from that as well.

Many of our social ills seemed to be connected to a lack of transparency,
and an undue focus on proforma privacy.  Institutions, priesthoods and the
like seem to hide and hold power and it feels important to democratize
participation, especially participation in community, not to encourage say
voting, or rallies but to encourage direct action.

Clearly one could build out any kind of social signaling network; one could
throw up a server, write some small iphone app, let people message each
other... but there an existing infrastructure now via twitter and like
services...  and so much of making ideas succeed seems to be minimization of
barriers, using existing infrastructure and the like.  Twitter has a
critical mass that is hard to beat.

So, in any case, thinking out loud this morning - and hoping for input and
direction - as a preamble to possibly presenting at wherecamp:

Citybot effectively creates groups within twitter.  It re-echoes all
statements that are sent to it - updating once a minute - effectively
creating a shared group where you can hear from voices that are not
necessarily voices you explicitly subscribed to.

One of the holes in twitter is that it requires everybody to explicitly know
everybody else - you can't "join a room" and "listen in on the conversation"
- instead you have to make explicit friendships with each of the persons in
a conversation one by one.  It means the act of finding people to talk to is
much aggrieved and a constant process of scouring - lest you only get half
the conversation.  As well you see pathologically large collections of
follows and followers attached to each person because it is not humanly
possible to scale this approach.  The concept of 'groups' answers this
problem - although they have their own problems - but groups are a concept
that has worked well for a long time, and follow an observation "the more
that service is like IRC the more successful it will be".

If twitter had groups then there would be no point to citybot in its current
incarnation.  Folks could listen to a room called "Portland" or I dunno
"Boston" and be done with it.  There is some merit in perhaps offering
automated room creation - in that you ask citybot to echo all messages not
to say the citybot channel but a specified channel for which it had login
permissions.  But likely twitter will add groups soon.

*

In any case emulating groups isn't really where I want to go with citybot.
What I want to try to do is to provide a new signalling network across
Portland (specifically I want to focus on my geography these days) that
connects people over complementary needs and helps rewire the city in a way
that ordinary folks can do better during the current economic fluctuations.

Like Craigslist I'd like to offer search services that don't just reinforce
existing social networks and friendships but help create new ones.  I admire
this about Craig's project.

Also, again like Craigslist, I really want to solve real problems for real
people; not just talking to a ditherati. Our social networks need to be
repaired - due to a legacy of an institutionalized capitalist infrastructure
that says every person is an island and should acquire goods and services
through dollar transactions and marketing and the like.  We trust walmart
more than we trust each other.

But unlike Craigslist I'm really keen on this being real time.  And keen on
better geographic information.  And would like least some kind of filtering;
not to the extreme that Twitter does it - but with at least a basic respect
that our social network graphs are too easily invaded by predators.  I'd
like to be able to have second and third degree orders of trust; so that if
messages were from folks who were entirely outside the social network that
they didn't get as much airtime as messages from people you trusted.  Like
for example if you had a cap of 100 messages per day - then the priority
should be on messages from folks you can trust.

So.

My feeling is that in its current incarnation Citybot doesn't provide huge
value yet.  But I see features I can do to get closer to what I want:

1) I can offer more groups - this is a low priority...  it is something I
could do.  I do feel a failure of groups is that they work too well and
people disappear into them instead of staying in a broader conversation...
I like to limit how many groups there are.  And I think twitter will solve
this soon.

2) I can actively matchmake complementary interests - there is no
matchmaking yet at all.  This I think is good and a good next step.  I would
like to have humans as matchmakers however rather than just robots. I can
imagine a lot of "cb junkies" that will hang out on the channels and connect
folks together - the same way that some folks hang out on cb radio channels
in small towns near truck stops and connect truckers to important
information about their landscapes.  There is a lot of the street watching
crowd that would help in a passive way here and might even get points for
it.

3) I could actively scavenge social friend networks by traversing the social
graphs of people that message citybot and use that as a basis for filtering
any active matchmaking that I do.  This would even work for corporate
messages and could help fix advertising.  I've earlier pointed out what I
call "the pizza problem" - finding a good pizza late at night.  If you could
signal that you wanted pizza then some corporation that had pizza that was
friended by your friends would be able to message you.  I also see this as
important in inverting the kind of google search model - where even if you
have the money you have to engage in the burden of search.  If you have
dollars, and you stick your hand up, the people who want your dollars should
be able to reach you in a filtered way - you should not have to carry the
burden of search.

4) I think there has to be a marketing and promotional effort if this kind
of project is going to succeed.  Many projects which at least I thought were
quite good ( such as http://thingster.org ) were failures.  I suspect better
outreach and participation is needed.

5) Privacy support - it would be good to have this because often people want
to signal each other discreetly for things that are socially unacceptable or
the like.  In the early stages of delicious Joshua decided against privacy
(for delicious) - and I agree with that - it was the right thing to do -
privacy slows the growth of a service because it hides its activity.  As a
later feature it may have value.

The real question is - what to do first - what is best - what is most
critical?  Can this idea succeed?  What does it take?  How can I build
support for it?  I'd like to find ways to make my communities work better;
it feels like this is in the right direction.

   - anselm 415 215 4856 http://hook.org http://makerlab.com
http://meedan.net
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking

Reply via email to