Subject: RE: The Internet, Skepticism, and Self-Perpetuating Revolution |
From: "Yasha Hartberg" <yasha.hartberg@gmail.com> |
Date: 3/21/10, 14:58 |
To: "'Barrett Brown'" <barriticus@gmail.com> |
Barrett,
I look forward to seeing your other writing on the
subject. I should probably emphasize that I am optimistic that your
network can improve the flow of information and that it may effect positive change
in at least some domains. Nevertheless, I would predict that there will
be other domains that will prove as stubborn as ever. Mapping out these
different domains would, in fact, be fascinating to me. I imagine that
the way you are setting up this network should make it relatively easy to track
information flow. I am curious, though, what sorts of metrics you may
have in place to assess whether that improved information flow has the impacts
you desire.
Thanks for your time,
Yasha
From: Barrett Brown
[mailto:barriticus@gmail.com]
Sent: Sunday, March 21, 2010 10:54 AM
To: Yasha Hartberg
Subject: Re: The Internet, Skepticism, and Self-Perpetuating Revolution
Yasha-
That's fantastic that you're studying with Wilson. I hope
I'll be able to convince you of the relative importance of information flow, if
not by way of these materials, than with some of my other writings on the
subject.
This
effort is called Project PM. The major goals are to (a) discredit prominent
pundits such as Thomas Friedman and Charles Krauthammer in order to reduce
their negative influence on U.S. policy and (b) increase the positive influence
of the more capable segments of the blogosphere. Both of these goals are
dependent on the deliberate manufacturing of critical mass among bloggers such
that segments of the traditional media will be prompted/forced to address
certain critical issues as determined by a collective array of a network of the
most erudite bloggers in existence. A third goal that does not require the same
critical mass or temporary control over the traditional media infrastructure
involves the development of a communicational schematic that is superior to
anything else in existence in terms providing bloggers with the best possible
feed of raw information by which to produce content, as well as the best means
by which readers can most easily find the best and most important of this
content without having to sift through duplicate or sub-standard info, which is
to say that it will discard some of the problems inherent to reddit,
memeorandum, and other such sources as exist today. All of this is to be
achieved by way of a network I've designed, and which will make use of
open-source software being designed for the purpose and developed by a fellow
with a brilliant track record in innovative IT implementations.
Below,
I've pasted a rough draft summary of the network and how it functions. Take a
look and let me know if you have any other questions. Feel free to forward this
to anyone else that you consider to be intellectually honest, capable, and
interested in doing something ambitious.
Thanks,
Barrett
Brown
Brooklyn,
NY
512-560-2302
Project
PM Network Summary
The institutions and structures that have developed over the past two decades
of accelerating public internet use have had what we reasonably describe as a
wholesome effect on information flow. But the information age is a work in
progress, and thus there are potential improvements to be made. More
importantly, there are improvements that can be made by an
initially small number of influential participants working in coordination. The
purpose of Project PM is to implement these solutions to the extent that
participants are collectively able to do so, as well as to demonstrate the
beneficial effects of these solutions to others that they might be spurred to
recreate or even build upon them independently of our own efforts.
The Problems
Project PM is intended to address the
following inefficiencies:
(a)
Watering down of contributor quality within participatory networks: Open institutions
such as reddit.comtend to peak in terms of the erudition
of the content conveyed a few years after coming about, with this being due to
the particular dynamics of network growth. By definition, early users are early
adapters, who themselves tend to be better-informed and otherwise relatively
capable in terms of the value they bring to the network. To even know of such
networks early in their existence is to pass a certain sort of test regarding
the potential quality of one's contributions; as knowledge of the network expands,
this "test" becomes easier, and to the extent that it does, the
network is less "protected" from those who did not pass such a test
by virtue of the fact that they did not know of the network until knowledge
became more common. Obviously, failing to be aware of some particular
institution does not come anywhere near precluding one from being intelligent
and knowledgable in general and thus of value to the institution, but the
influx of valuable participants versus damaging participants appears to decrease
after a certain level of notoriety is reached. Again, the decline in the
intellectual relevance of content at reddit.com is a good
example of this.
(b)
Data overflow: The
watering down process described above does not only result in one coming across
information of relatively low quality, but also in having to contend with more
of it. On reddit.com, for instance, a user who scans new
submissions will find not only a certain amount of potentially useful
information, but also some amount of almost certainly useless information. The
watering down of contributor quality also contributes to the extent to which
the latter is perpetuated within the network itself insomuch as that lesser
contributors are more likely to vote up useless information, thus helping to
ensure that the barriers built into the network in order to facilitate the
viewing of important rather than unimportant content - in this case, a
pre-established threshold of up votes necessary to bring something to the front
page - will thereby lose their effectiveness.
(c)
Barriers to obtaining raw data: The obvious fact of data overflow - that
some data is more useful than other data - is dealt with by means of selecting
certain sources of information which one has identified as being a provider of
quality output relative to other sources. Bloggers and others who require a
steady stream of data in order to operate have certain methods of obtaining
that data, and there is of course no reason to believe that any of these
methods could not be improved upon to an extent that these improvements would
be worth adapting. One has RSS feeds flowing from sources one has selected (and
by virtue of having been selected, the sources must have been necessarily known
to the blogger in the first place); one has algorithm-based sites like
Memorandum.com (which merely shows what bloggers are talking about rather than
necessarily providing any insight into what they should be talking about); one
has democratic or pseudo-democratic sites such as reddit.com and digg.com;
and one has the fundamentally one-way outlets of television and newspapers, the
content of which is decided upon by a handful of producers or editors (who
themselves are working within an incidental structure that does not appear to
be of much value relative to what may now be found among the better portions of
the blogosphere). A means of obtaining data that improves upon these and all
other methods would be of great utility insomuch as that the quality of data is
of course one major limiting factor with regards to the quality of output..
The
Solutions
By way of a network designed to take better advantage of the existing
informational environment, Project PM can help to remedy the problems described
above without significant effort on the part of participants, yet with
potentially dramatic results on the efficiency of information flow.
(a)
Watering down of contributor quality within participatory networks: Project PM will
greatly reduce the accumulation of low-value contributors by way of the method
by which contributors are brought it. The network will be established with a
handful of contributors who have been selected by virtue of intellectual
honesty, proven expertise in certain topics, and journalistic competence in
general. Each of these contributors has the option of inviting into the network
any number of other bloggers, each of whom will initially be connected only to
the contributor who brought him in. Each of these new participants also has the
option of bringing others into the network in the same fashion as well as
offering a connection to any other participant, as will anyone they bring in,
and so on. To the extent that the original participants are of value in terms
of their judgement, they may be expected to bring in participants of similarly
high value, and so on; meanwhile, as the network expands, participants will be
likely to form new direct connections to others whom they have determined to be
of particular value relative to other participants, and conversely, to
disestablish any direct connections they might have established to those whose
output they find to be below par. Of course, none of this precludes the network
from eventually encompassing participants of low desirability relative to that
of the average participant, but to the extent that such a thing occurs, its
effect are largely neutralized by way of the dynamic described below.
(b)
Data overflow: Information
flows through the Project PM network by way of a single button accessible to
each participant. When a participant either writes or receives a blog post or
other informational element, the participant may "push" the item,
thus sending it to all of those with whom he is directly connected in the
network. In such a case as a participant pushes forward items that others may
determine to be of little merit, the resulting clutter is only seen by the
participant who brought such a low-value blogger into the network in the first
place, as well as those whom the low-value blogger has to this point brought in
himself along with those who have agreed to connect with him from elsewhere in
the network. To the extent that a given participant exercises good judgment in
establishing connections, then, he will only receive informational elements of
value while also being able to quickly transmit them to contributors who will
be able to make best use of such information. Meanwhile, below-average
participants will have only very limited means by which to clutter the network,
as informational elements become less likely to be pushed forward as they
approach above-average participants within the network, who themselves are
"buffered" from such things by way of the competent participants with
whom they surround themselves by way of their connections and who, by virtue of
their competence, are unlikely to push forward low-value information.
(c)
Barriers to obtaining raw data: The dynamics described in (a) and (b) collectively
provide for a means of information inflow that should theoretically be superior
to any other medium currently in existence in terms of overall quality, both by
virtue of the network's improved organizational methods as well as the
relatively high competence of participating bloggers relative to members of the
traditional media outlets as a whole. Accessibility to particularly valuable
items of information will be enhanced further by the option to set one's widget
in such a way as to display any piece of information from the network,
regardless of "proximity," if such information is pushed forward
(which is to say, approved of other participants) a certain number of times.
This should help to ensure that, as the network expands, particularly valuable
information does not become unduly "regionalized." A variant on the
widget for use by readers (as opposed to network participants) displaying
information that meets similar thresholds of popularity within the network
would likewise provide those readers with a source of information above and
beyond other existing mediums.
On Sun, Mar 21, 2010 at 9:45 AM, Yasha Hartberg <yasha.hartberg@gmail.com> wrote:
Greetings!
I
came across your blog posting on True/Slant this morning and thought that your
interests may intersect with my own. I am a second year graduate student
at Binghamton University studying cultural evolution with David Sloan
Wilson. In particular, I am interested in understanding why cultures are
resistant to change. I am still in the early stages of forming an
evolutionary theory for why we should expect cultural recalcitrance in
certain contexts. Nevertheless, I have collected enough background
material to suspect that improving information flow as you suggest may have
only a limited effect.
I
would welcome a dialog with you about these topics.
Sincerely,
Yasha
Hartberg
Department
of Biology
Binghamton
University