Inkwell: Authors and Artists

    
I don't know how to fix the structure of democracy. I have some degree
of hope and enthusiasm for efforts to connect citizens to citizens,
and to connect those in charge to those citizens. There have been lots
of different attempts, some more successful than others. But all of
them aim at affecting the processes that lead to decisions, rather than
at how decisions are themselves made.

I talk in 2b2k ("Too Big to Know," in case that's not clear) about the
effect of the networking of knowledge on business decision-making. It
was a particularly difficult chapter to write, and probably one of the
least convincing ones. It makes a couple of points that I think are
right, but I didn't (and don't) know how to make them go far enough
that I could claim (as I and my publisher would have wanted to) that
the networking of knowledge transforms decision-making within the
business world. Instead, I try to get close by refocusing "decisions"
away from that moment that the boss goes thumbs up or down. Instead, we
can see the effect of knowledge networks if we think about decisions
within the context that led up to them and the processes by which those
decisions are appropriated by the business. 

I also suggest that businesses have a lot to learn from the large
collaborative Web projects that have created and maintained products of
a sort that formerly had been produced through hierarchical systems,
e.g., Wikipedia, Linux, etc. These collaborative enterprises tend to
keep decision-making as local as possible, because that's where the
best knowledge is. The alternative is that you move info up the
pyramid, reducing it at each step, so your Jack Welch boss-man can make
a decision. But, I acknowledge in the book that there are legal as
well as cultural reasons why most businesses won't soon move to
non-hierarchical, collaborative decision-making. And it'll take
governments even longer since they tend to have those pesky
constitutions.
  

    
> These collaborative enterprises tend to
> keep decision-making as local as possible, because that's where the
> best knowledge is.

We even might say that the most effective knowledge is that which is
situated within the problem domain itself.
  

    
> The alternative is that you move info up the
> pyramid, reducing it at each step,

A possible varient of this (which the Net makes increasingly
practical), is to *expand* the information (dare we say 'knowledge?')
available at each step.

In the past, reduction was necessary to accomodate the bandwidth
limitations of each higher level. But the Net (and it's implicit
computational abilities) can drasticlly transform these limits.
  

    
Craig, we might indeed say that (i.i. knowledge  situated within the
problem domain).

As for expanding the knowledge at each step: On the one hand, I agree.
2b2k is about how the Net is enabling us to move from a reductive
knowledge strategy to an inclusive one. But, even if you change the
pyramid to something more like a rectangle, so that we don't have the
implication that each step in value is a reduction in size, I'm not
happy with an architectural diagram that suggests that wisdom is built
on knowledge that is built on info that is based on data. I think the
relationships among those four terms are far far more complex than
that, and don't lend themselves into a stack order.
  

    
I completely agree. It makes some sense to me that DIKW arose when it
did -- as a conceptual advance that addressed some real problems of
the time. But today, it's a legacy notion.

It also seems to me that DIKW came about when the efforts of professional
philosophers was not yielding practical concepts of knowledge that could
be applied (and worked) 'in the field.'

This leads me to ask: do you think there is any branch of philosophy
today that might ultimately become the source of a more workable and
precise theory of knowledge? Or has it become an engineering exercise?

I would imagine that the myriad ways in which are seeing knowledge
behave in various networks (ie. 'in the wild') would become the source
of the field observations used to perfect a new theory of knowledge.
  

    
Reading the comments about decision-making -- whether in business or in
a democracy -- in the context of Michael Lewis's profile of Barack Obama
makes me skeptical. As Lewis points out, Obama faces immediate opposition
to most decisions that he makes, whether it's from the Republicans
(who oppose him and his party) or Russia (whose leaders oppose the US).

The same dynamic happens in a business context. Bosses, from boards to
CEOs to lower managers, make decisions based on personal or political
motivations all the time, regardless of the information available. Which
is to say that all of the information is NOT available, because you
can never achieve complete transparency of all the factors informing a
particular issue. It can be as simple as, "I can't publicly agree with
Bob's premise, because that will give Bob leverage against Charles,
who is my ally. So I must, regardless of the merit of his information,
oppose him."

So, in the case of arguing the merits of the Affordable Care Act,
no matter how detailed the conversation gets, the availability of the
information in any quality or quantity may never overwhelm the emotional
opposition to the source of the information. False information gets
created to sway decisions based on emotions rather than "truth." (I
suppose this may not even be a conscious process.)
  

    
I've had the opportunity to operate in supposedly more democratic
business contexts, and saw that the flatter hierarchy could work well
or could create toxic environments. Culture and individual mindset are
key:  if you have a culture of cooperation with strong buy-in among
members of the organization or team, a democratic approach can work
well. If you have members who prioritize their personal desires and
ambitions and try to build a power base and game the system to their
own ends, the environment can grow toxic - more so if you have multiple
members like that. Our recent guest Bruce Schneier wrote in his book
_Liars and Outliers_ about how trust and cooperation hold a society
together, and how we overcome self-interest by realizing that our
individual needs are better served by cooperation. The difficulty of
getting everybody on the same page about this is why we have security
systems, and also why we have vertical hierarchies of authority,
police, courts, prisons etc. - because there are inherently and
persistently "defectors."

This may seem irrelevant to a discussion of knowledge, but I think
it's relevant to the discussion of authority, which is an important
aspect of knowledge. (By what authority do we decide that a fact is
"true," or a collection of facts and assumptions can be committed to a
shared body of knowledge?)  I think "belief" is in the mix, as well,
and we haven't talked about that yet - does knowledge require "faith"?
  

    
Scott, I agree with your description of many or even most environments
in which business decisions are made. But I'm not sure what exactly
you're skeptical about in your first sentence. 

(BTW, I posted briefly about the Lewis article yesterday:
http://www.hyperorg.com/blogger/2012/09/16/2b2k-decisions-and-character/)
  

    
Jon, you write: "(By what authority do we decide that a fact is
"true," or a collection of facts and assumptions can be committed to a
shared body of knowledge?)"

Nice committing the hub of two generations of thought to a mere
parenthesis, muh friend. Fortunately, I have an equally pithy answer:
"What you mean _we_, white man?" [Note: This is a Mad Magazine
reference, not a random racist remark.] 

We used to think that we decided which facts were true, but then it
became obvious that those decisions were generally being made by the
privileged class that owned the printing presses and broadcast towers. 

We're now in a mixed up stage, which may turn out to be the permanent
normality, of recognizing, resisting, and forgetting about the social
nature of knowledge authority, pretty much simultaneously. This is,
after all, at the heart of the echo chamber problem looked at
epistemologically: when you're in an echo chamber, you forget that the
views expressed have authority only within it. (The echo chamber
problem has a different heart if you look at it sociologically or
psycho-neurologically.) Likewise, even the revered Wikipedia is only
possible because of some commitments to what sort of discourse lets a
contributor become part of the "we," and as many have pointed out (h/t
to Joseph Reagle, among others), the old gender biases reappear there.
(Wikipedians are well aware of this and struggle with the problem.)

This leaves us with a muddle. On the one hand, we want to take
seriously the social nature of truth, with both its good and bad
effects. On the other, I by no means want to say that all truths from
all the world's we's are equal. I just don't know how to distinguish
one from another except from within my own group, which is exactly the
problem. 

So, we muddle through, in contention.
  

    
I'm skeptical of the thoughts I perhaps misinterpeted in #25 and #26: that
our emergent "tools for conversation" will somehow transform the process
of making decisions, in a business or in a democracy, because (I believe)
the "facts" will never be completely objectively stated or absorbed.

Flat and egalitarian a system may be, but I'm not sure that the increased
exposure to information or a greater number of voices will actually
change the emotional context in which people converse. I'm not at all
arguing that these new tools are not beneficial, just that I'm skeptical
anyone has become a more rational actor because of them. I'd like to
believe that, but I don't.
  

    
Slip, somewhat underscoring my thoughts.
  

    
Thanks for the clarification, Scott. 

I certainly agree that relying on objectively stated facts to be
agreed upon and absorbed is not going to get us further than we
currently are when it comes to making decisions. I like facts (as do
you), but decisions inevitably are not and cannot be purely fact-based.
(I have a chapter in 2b2k about the history of facts and of this
particular hope.)

But I do think that in some environments, network-based decisions can
help. That's not because they have a better fact-base, but because they
can leave many decisions in the hands of local subnets that know more
and understand better what the local situation calls for. (Note that
this understanding is not purely factual.) When decisions need to be
scaled up -- not just handled at the local level -- networks also
provide some benefits. In particular, they include more people arguing,
rather than reducing the amount of info as it goes up the pyramid.
  

    
Aside from these issues of authority and knowledge, in thinking about
democracy we also have the problem of scale, which is also relevant to
network knowledge. How do you scale a conversation?  This conversation,
with the two of us and some few others who, like <esau>, join and
contribute, is quite manageable. However if we carried this
conversation to a high-traffic public context, and tens or hundreds of
contributors chimed in, we couldn't keep track. If you had a hundred
questions popping up every day, how would you answer? 

Some items on Reddit, for instance, have hundreds of responses and
dozens of threads. It could take all day to thoroughly read just one or
two items.

Another example: I once joined a popular Twitter chat for bloggers and
set the Twitter stream for the conversation to update more or less
realtime. The flow of responses reminded me of a slot machine. I
couldn't even read them, let alone respond?

This seems like a wicked problem.
  

    
Yeah, during the Democratic National Convention, I watched the stream
of tweets with the #dnc2012 hashtag, and you could practically hear the
whoosh as they went by.

How do you scale conversation? All the ways that we've invented, each
with their own strengths and weaknesses. Reddit scales conversations in
one way, Facebook in another, The Well in yet another. And I'm sure
we'll come up with many, many more.

If the question is how to scale conversation so as to increase
knowledge, and to make knowledge networks smarter, I'm unfortunately
going to have the same type of answer. I'd still point to The Well and
to Reddit, but FB would not have sprung quite so quickly to mind. But
there are lots of other good answers to this question.
  

    
Do you think the quality of conversation is lower on Facebook? I tend
to call Facebook and Twitter sources of "drive-by" conversations, short
bursts that are relatively shallow. However I've seen conversations
with more depth on Facebook from time to time. However Reddit seems to
be the best place for deeper conversations today, and I assume there
are others. The Consumerist is also a good example - a small staff of
bloggers post stories of consumer interest, and a robust community of
readers respond with great, often snarky discussions.

Where do you, personally, look for conversation online?
  

    
I like Reddit very much, although probably about 95% of the topics on
the front page I have no interest in, I don't understand, or are worth
only a quick check. Still, that means that I find one or two topics
worth at least browsing every time I check. I like the depth of
expertise, the pleasure they take in altruism, the nerd-geek
conversational aesthetic, and the mixing of the insightful with the
silly. I also am fascinated by IAMA's as a new journalistic form, and
by the highly developed collaborative humor shown in many of the
threads.

But that's me. As is true of all of us, some forms of conversation
suit me better than others. I am not very good in public threads; I
find them intimidating. I'm more relaxed in mailing lists where I've
been a member for a while. I like tweeting, but I have trouble using it
for conversation because of the asymmetry: the people reading my
response are not necessarily the people who read the tweet I'm
responding to.

But I would not say that I am a particularly good example of someone
who converses well online.
  

    
For those who don't know Reddit, "IAMA" is a subreddit, which is kind
of like the conferences here on the WELL - a broad subject area that
includes a lot of conversations. In that particular subreddit, famous
people or people who have interesting backgrounds or jobs join an "AMA"
("ask me anything") conversation. The most visible of these recently
was Barack Obama's:
http://www.reddit.com/r/IAmA/comments/z1c9z/i_am_barack_obama_president_of_the
_united_states/sort=new

Your mention of threads reminded me of the threadsml project you were
involved with some years ago - as I recall, it was an attempt to come
up with a standard for threaded conversations so that they could be
shared across platforms. Whatever happened with that project?
  

    
David, I'm wondering what your response is to this article by Sandy
Pentland on the Edge?:

http://www.edge.org/conversation/reinventing-society-in-the-wake-of-big-data

How much do the algorithms predetermine data capture and possibly
affect these societal changes?
  

    
One thing I know is that it is my responsibility to protect both my
identity and my data from here on in. I suppose that's part of the
process of any emerging technological revolution. Having been on the
Net for 20 years I'm finding it more difficult than ever. How do see
personal data protection evolving in the future? I would hope the
"room" would be a bit more friendly than it is now.
  

    
(Sorry. I was on trains almost all day yesterday, with poor wifi.)

Jon, good memory! Threadsml was just as you describe. Except that we
didn't get anywhere with it. I still think it'd be great to have a good
way to get comment boards to interoperate.
  

    
Ted, thanks for the pointer to the Sandy Pentland article.
Interesting. I think it'd push back on his denigration of experimental
work in the face of Big Data. BD is hugely exciting, but there's lots
of room for experimental science, as I think Pentland would agree. 

The comments most relevant to our discussion at the moment, though,
have to do with his understanding of understandings. Here's a
paragraph:

"The other problem with Big Data is human understanding. When you find
a connection that works, you'd like to be able to use it to build new
systems, and that requires having human understanding of the
connection. The managers and the owners have to understand what this
new connection means. There needs to be a dialogue between our human
intuition and the Big Data statistics, and that's not something that's
built into most of our management systems today. Our managers have
little concept of how to use big data analytics, what they mean, and
what to believe."

With complex systems, we well may not be able to understand what's
happening or why it's happening. And our intuitions are likely to be
wrong. 

I do of course like his proselytizing for organizations with more data
sharing and transparency.  
  

    
Ted, about privacy: Something is off with my cognitive privacy
assessment subsystem. I am personally toward the edge of the privacy
scale for much about my life, but much less so on the Net. Worse, I
have no enthusiasm about privacy laws and policies. I seem just not to
care much, probably because no one has hacked my phone and posted all
those salacious photos I keep there, and no one has targeted any of my
children for online abuse. I know I'm supposed to care about privacy,
and I'm glad other people do care, but I just sort of don't.
  

    
I've come to the point where I don't care that much about my
'identity'. I'll control and shape that in my own way and hope for the
best. I am concerned about my data tho, who owns it, who uses it, etc.
That's a much bigger issue for me as a 'data point(s)' and I'm not
happy with all these silos and walled gardens.

The quote cited from Sandy's piece ... is that a good touch stone for
what's developing in the sense of new data and knowledge networks as
you talk about them in your book?
  

    
Btw, this observation:

> With complex systems, we well may not be able to understand what's
> happening or why it's happening. And our intuitions are likely to be
> wrong.

is an argument for data-privacy, isn't it?
  

    
Ted, as you know, the problem is that if you want to break down the
walled gardens (and don't we all!), then you also make it far easier
for your personal info to be spread and abused. I don't know what to do
about that tension.

The passage I cite from the Pentland article points to one of the
characteristics of networked knowledge, but I'm not fully in agreement
with what he makes of it. So, we agree that Big Data and Big Knowledge
lead us into a world that surpasseth human understanding. Sandy
recommends using our intuition. But human intuition scares the bejeebus
out of me. I think we're just stuck with knowledge without
understanding, especially as knowledge becomes networked. 

I'm not being very clear (how ironic!) so let me unpack it a bit.

We've always had knowledge without understanding. That happens at the
beginning of the process when we notice a correlation but don't
understand why it occurs. But more importantly, it also occurs at the
other end, when the knowledge becomes commoditized. We all know all
sorts of things we don't understand, and I'd venture to say that there
is nothing any of us knows that we fully understand. E.g., I know about
gravity, but do I understand it? Nah, not if I think about it for more
than 2 seconds. We don't have enough brain capacity to know only that
which we fully understand. This has been a prime element in our
traditional strategy of knowing the world by reducing it.

We've been able to reduce knowledge because we have media that
preserve it and make it accessible beyond the individual.
Traditionally, that's been paper, books, and libraries. But those media
are very limited. Now we have a medium with indefinite capacity. We
don't have to throw out a Web page to make room for a new one, as
libraries eventually have to do. (Offsite book repositories are one
response to this.) Plus, of course, our new medium is linked with linky
links out its linky-linked wazoo.

So we still have lots we know that we don't understand, but now
there's much more of it. Knowledge networks give us knowledge that is
understood only at the level of the network, by a community that is
linked and shifting. Big Data gives us knowledge mediated by computers
that contain and process more than our brains can manage.

Too Big to Know maintains that this change is not just a change in the
quantity of non-understood knowledge but is also a change in
knowledge's nature. But it's important to recognize the continuity as
well as the disjunction. Nothing is ever truly new, as we have known
forever.