Smartocracy - a tool for meritocratic collective decision-making
23 views
Skip to first unread message
Brad deGraf
Oct 19, 2011, 1:19:07 PM10/19/11
to OWS Digital Strategy
I wanted to offer a collective decision-making tool that is ideally
suited to OWS. It was originally developed for foundations to
distribute their funding decisions to large networks of experts. The
decision-making engine could easily be applied to many kinds of
decisions from mundane daily ones to momentous global ones. It scales
to millions of users and could easily be integrated with Facebook
Apps.
My company Sociative is using it to filter the noise out of Twitter
(we're applying it to OWS now and will make that public asap).
If you're interested, you can see some documentation at
http://sociative.net/?page_id=204
suited to OWS. It was originally developed for foundations to
distribute their funding decisions to large networks of experts. The
decision-making engine could easily be applied to many kinds of
decisions from mundane daily ones to momentous global ones. It scales
to millions of users and could easily be integrated with Facebook
Apps.
My company Sociative is using it to filter the noise out of Twitter
(we're applying it to OWS now and will make that public asap).
If you're interested, you can see some documentation at
http://sociative.net/?page_id=204
Charles Lenchner
Oct 19, 2011, 1:51:34 PM10/19/11
to ows-digita...@googlegroups.com
I'd love to see a demo of this.
I'd be cautious about offering something that is branded/not open source.
I'd be cautious about offering something that is branded/not open source.
Thanks!
Charles
Chris Hoffman
Oct 19, 2011, 1:54:18 PM10/19/11
to ows-digita...@googlegroups.com
Brad deGraf
Oct 20, 2011, 10:36:46 AM10/20/11
to OWS Digital Strategy
Totally understand and expect the desire for open source. Let's
explore it. We're looking in to what we can open-source practically.
Would black-boxing the engine work? For instance, could we provide the
engine along with a test harness that anyone can run against it to
ensure expected/reasonable results?
Doesn't need to be branded. We just really believe that it would
seriously facilitate rapid action on many levels and would love to
contribute any way we can that doesn't hurt our business.
Brad
On Oct 19, 10:51 am, "Charles Lenchner" <clench...@organizing20.org>
wrote:
explore it. We're looking in to what we can open-source practically.
Would black-boxing the engine work? For instance, could we provide the
engine along with a test harness that anyone can run against it to
ensure expected/reasonable results?
Doesn't need to be branded. We just really believe that it would
seriously facilitate rapid action on many levels and would love to
contribute any way we can that doesn't hurt our business.
Brad
On Oct 19, 10:51 am, "Charles Lenchner" <clench...@organizing20.org>
wrote:
Brad deGraf
Oct 20, 2011, 10:39:45 AM10/20/11
to OWS Digital Strategy
also, see http://sci.r88r.net for a Science 'zine that uses
smartocratic networks of experts. Click on 200 influencers and mouse
over names to get an idea of how it works.
smartocratic networks of experts. Click on 200 influencers and mouse
over names to get an idea of how it works.
Charles Lenchner
Oct 20, 2011, 10:42:54 AM10/20/11
to ows-digita...@googlegroups.com
Brad,
Push back the concerns over OS and branding. Most important thing is making
the value clearer.
There are multiple decision making tools on the market; help us understand
your feature set better in comparison with the alternatives. There's a lot
of latent support for adopting such a tool, but folks need to understand
better why THIS tool is the right one.
Help us shop wisely for decision making software that addresses real needs
and likely use case scenarios.
Thanks
Brad deGraf
Oct 21, 2011, 12:11:36 PM10/21/11
to OWS Digital Strategy
ok, good questions. Here are the advantages of the smartocratic
approach:
1. Meritocratic - A participant's weight in a decision is a function
of their 'people rank' in the network of voters. Each voter has an
equal amount of 'proxy' to delegate to other participants (edges in
the social graph of the network are delegations). Those people who are
delegated to more have more 'mojo'. A delegation from someone with
high mojo is worth more than one from someone with low mojo. It is
thus iterative in much the same way as Page Rank.
2. Optimally Participatory - No participant's votes are wasted if they
don't/can't participate. Their weight automatically flows to their
delegates, until reaching one or more who do vote. One major benefit
of this is that it encourages a voter to abstain if they don't feel
well-informed, because someone better informed will represent them.
3. Contextual - Delegations are context-dependent. I might want to
delegate to one set of people for finance but another for music or
whatever. Effectively, this means a different influence network per
topic.
4. Iterative and Near-real-time - Votes don't have to be atomic. The
system is constantly re-ranking based on new delegations and votes, so
that one can view the current state of the decision all the time (if
desired, not necessarily). So participants can see where it's going,
dig deeper based on that, and maybe change their thinking.
5. Bootstrapped - We already use Twitter attention (RTs, @refs, lists,
follows) as delegations, and thus have working networks for literally
thousands of topics and can make new ones easily. For instance, I can
provided a picture of one made from 2168 twitter users who used the
#ows hashtag two days ago. (how do I attach a file here?)
[CONFIDENTIAL: you can view networks at our currently private
aggregator site http://occupy.r88r.net/ PLEASE DO NOT FORWARD THAT
LINK YET. IT'S NOT READY. click on 200 influencers and mouse over
names. Most meaningful ones are the Raw Feeds. The hashtag one
mentioned is 'hashtaggers']
6. Social Media friendly - Besides the bootstrapping from existing
behavior, delegations could easily be made with a tweet, e.g. "#ows
#delegate @foo #topic" could create a delegation from the sender to
user foo on the topic 'topic'. It is also quite easy to make a
Facebook app that takes a group of users and their votes among a set
of candidate choices and yields a current result based on those.
7. Scaleable - The system is architected to scale to millions, and to
nest (e.g. one could have a national network composed of local
networks).
8. Mature - We did version 1 over 6 years ago, and this version
started 3 years ago. It is in heavy use pulling signal out of Twitter
Noise.
I'll leave it there. Looking forward to your thoughts.
Brad
On Oct 20, 7:42 am, "Charles Lenchner" <clench...@organizing20.org>
approach:
1. Meritocratic - A participant's weight in a decision is a function
of their 'people rank' in the network of voters. Each voter has an
equal amount of 'proxy' to delegate to other participants (edges in
the social graph of the network are delegations). Those people who are
delegated to more have more 'mojo'. A delegation from someone with
high mojo is worth more than one from someone with low mojo. It is
thus iterative in much the same way as Page Rank.
2. Optimally Participatory - No participant's votes are wasted if they
don't/can't participate. Their weight automatically flows to their
delegates, until reaching one or more who do vote. One major benefit
of this is that it encourages a voter to abstain if they don't feel
well-informed, because someone better informed will represent them.
3. Contextual - Delegations are context-dependent. I might want to
delegate to one set of people for finance but another for music or
whatever. Effectively, this means a different influence network per
topic.
4. Iterative and Near-real-time - Votes don't have to be atomic. The
system is constantly re-ranking based on new delegations and votes, so
that one can view the current state of the decision all the time (if
desired, not necessarily). So participants can see where it's going,
dig deeper based on that, and maybe change their thinking.
5. Bootstrapped - We already use Twitter attention (RTs, @refs, lists,
follows) as delegations, and thus have working networks for literally
thousands of topics and can make new ones easily. For instance, I can
provided a picture of one made from 2168 twitter users who used the
#ows hashtag two days ago. (how do I attach a file here?)
[CONFIDENTIAL: you can view networks at our currently private
aggregator site http://occupy.r88r.net/ PLEASE DO NOT FORWARD THAT
LINK YET. IT'S NOT READY. click on 200 influencers and mouse over
names. Most meaningful ones are the Raw Feeds. The hashtag one
mentioned is 'hashtaggers']
6. Social Media friendly - Besides the bootstrapping from existing
behavior, delegations could easily be made with a tweet, e.g. "#ows
#delegate @foo #topic" could create a delegation from the sender to
user foo on the topic 'topic'. It is also quite easy to make a
Facebook app that takes a group of users and their votes among a set
of candidate choices and yields a current result based on those.
7. Scaleable - The system is architected to scale to millions, and to
nest (e.g. one could have a national network composed of local
networks).
8. Mature - We did version 1 over 6 years ago, and this version
started 3 years ago. It is in heavy use pulling signal out of Twitter
Noise.
I'll leave it there. Looking forward to your thoughts.
Brad
On Oct 20, 7:42 am, "Charles Lenchner" <clench...@organizing20.org>
Brad deGraf
Oct 21, 2011, 12:17:14 PM10/21/11
to OWS Digital Strategy
a screenshot of a network is at
http://sociative.net/wp-content/uploads/2011/10/Screen-shot-2011-10-21-at-8.48.42-AM.png
In that view, AnonyOps is moused-over, showing their profile, and
their 'ins' and 'outs' (yellow is linked to them, red is linked from
them, both is mutual). In this case, those edges are Twitter attention
edges, which can be interpreted as delegation (of attention).
> aggregator sitehttp://occupy.r88r.net/PLEASE DO NOT FORWARD THAT
http://sociative.net/wp-content/uploads/2011/10/Screen-shot-2011-10-21-at-8.48.42-AM.png
In that view, AnonyOps is moused-over, showing their profile, and
their 'ins' and 'outs' (yellow is linked to them, red is linked from
them, both is mutual). In this case, those edges are Twitter attention
edges, which can be interpreted as delegation (of attention).
Reply all
Reply to author
Forward
0 new messages