Re: [Ag Meinungsfindungstool] Abschied

[Scott Raney <scott AT metacard.com>]

On Wed, Aug 27, 2014 at 9:56 AM, Martin Stolze <pirate.martin AT stolze.cc> 
wrote:
> Hi Scott,
> I added Altemeyer's book to my reading list but won't get around to study it
> anytime soon. As it stands, I can sense where you are coming from and I must
> disappoint you because you expend yourself on the wrong question.

Says the person without the background knowledge I'm using to guide my
inquiries ;-)

> To me it is a flawed approach to "design systems that that help people make
> rational decisions".

This not exactly the right way to phrase it, unless by "people" you
mean "The People" rather than any particular individual.  I'm
concerned with the former whereas I see you and the qKonsens project
focusing on the latter.  Individuals don't have to make rational
decisions so long as *collectively* the right decision is made.

> All of our decisions are exactly as rational as they can possibly be.
> Everybody is equipped with reason and enhancing it is a matter of
> education/learning.

Agreed so far.

> i.e. Coursera does that, everything else just fosters "faith/prejudice".
> Focusing on enhancing learning is a worthy pursuit!
> The problem on this end is that it requires "Investing enormous (and as a
> practical matter, unavailable) resources".

Agreed, again.

> On your list you forgot 4) Separate the different rationalities

This sounds like an aspect of 3) to me (correcting for irrationality)

> "what is the perspective from which we assess "perceived" utility?"
> I think Microeconomics has you covered already. Do you like the blue or the
> green T-Shirt? - The particular utility is inherent to each agent.
> We can however compare in that we ask: Would you like one or two T-Shirts?
> If you add opportunity cost you can determine the marginal utility.

Roughly, yes.  In practice I think this is a lot harder to do than
utilitarians account for, though.

> Another fundamental flaw: "but I believe"
> With that you set yourself up for failure and construct an ineffective but
> justifiable ideology: "All we need to do is recognize that irrationality
> exists"
> Yes we can, allahu akbar!

Whether it's ineffective or not is an empirical question, not a
philosophical one.  Which means neither you nor I can make this claim
until we actually start building a system based on the philosophy and
see if it works.

> I could not find the example you mentioned but I am happy to take a look if
> you point it out. My suspicion is that "authoritarians" don’t identify
> themselves as such and hence their worldview is consistent. It is not
> irrational. Look again on your
> list:http://en.wikipedia.org/wiki/List_of_cognitive_biases.

Again, you've got to read the book to really understand this.  But I
would put ideology, including authoritarianism (although it is more
innate than most other ideologies), on that list of biases.

> History has countless examples of people who seemingly act against their own
> interest. It is only that you can't comprehend with your mindset their
> mindset, that's all.

Fair enough.  But again, irrelevant to the design of our new system:
We don't have to understand the inner workings to create a *system*
that works in spite of them.  Unless you're proposing that "acting
against your own interest" is a good thing and one that we need to
preserve.  In which case I disagree and think that our existing system
delivers on that well enough without changing it (conservatives voting
to reduce inheritance taxes being the best example of that, because it
raises their own taxes while providing absolutely no compensating
benefit to them).

> Not at all, this is not philosophy, we aren’t that far yet.
> Let me phrase this differently.
> 4=f(2,2) ?
> The degree of how {conscious a decision} OR {informed a choice} is =
> f(cognitive capacity, exposure to information) ?
>
> It is by definition an issue we need to resolve before we can do anything.

Why?  You're proposing we have to understand the inner workings of the
mind before we can design a system that functions well enough to meet
our Goals.   I'm sure that's wrong, and a few thousand years of the
history of moral codes backs me up: Those old codes have a lot of
flaws, but they did an awful lot with very little understanding of
human nature compared with what we know now.

> If you were to try you would probably recognize that we actually have a
> "system that
> functions well in the world as it exists."
> Very well in fact! So well that there is no other system that can
> out-compete it right now. North Korea tries, so does China or the Theocracy
> of your choice.

We could quibble about where the dividing line for "functions well"
is.  I see the current system as having a lot of flaws, some of which
put the survival of our species at risk.  But if you'd prefer to state
the goal as "functions better than our current system", I'd be OK with
that.

>> I don't think it's correct to assume that the authoritarians make
>> *random* decisions, just that they will sometimes *all* make decisions
>> that are not optimal (e.g., in Altemeyer's experiment those 25% of
>> authoritarians would vote to put all authoritarians in jail, or ban
>> them from the voting system!) optimality being assessed however you
>> want (their own well being, the well being of the community as a
>> whole, etc.)
>
> I don’t assume that "authoritarians make *random* decisions". I don't
> believe they exist. You do by insisting on vague "irrationality".

Maybe "vague", but there are actually several behaviors covered by
"irrationality" so it's hard to be precise in every context.  But for
the most part I don't really care about the precise definition, only
that the system functions properly regardless of whether the
"suboptimal" behavior is the result of lack of information, cognitive
biases, instincts overriding cognitive processing, or whatever.

As to whether authoritarians exist, it's a highly reproducible
finding.  Maybe not as popularly known as the difference between
liberal and conservative or as the diagnosis of sociopathy,
narcissism, autism, bipolar, etc., but at least as reliable a
diagnosis as any of those.

(snip)

> The optimal outcome of any decision on a micro level is maximum utility for
> the agent. This does not exclude death by suicide attack if the personal
> utility function is determined "by the well being of the community as a
> whole".

Right, but again, I'm not concerned with the micro level, only the
collective level.  Our new system doesn't have "fix" irrational or
suboptimal behavior of the individual so long as *collectively* it
provides more optimal decisions than our current system.  Which is
IMHO a pretty low hurdle.

> It appears inevitable that "believe" is followed by "hope". Otherwise you
> would not make unfounded claims like: Brave and upright "Neurotypicals" are
> victims of "authoritarian" cock blocks who prevent them from getting tail
> and keep la bella vita out of reach.
>
> This part is pretty tendentious but I thought the same once upon a time, I
> am sure you’ll get around it as well.

I think we all have a rough idea of what high SDO and RWA individuals
are. Most people probably just think of them as assholes without
making any finer distinction.  But as it turns out there *are*
differences in the various aspects and *large* differences in the
percentage of the population with each of these different aspects.  So
it's important to know this stuff before designing a system that works
with all kinds of assholes.

> Even if you had 99% authoritarians, all the power would rest within the 1%
> of swing voters because the rest cancels each other out. Looking at the
> margin of the last 100 years in the US, it appears to me like a pretty
> perfect distribution.

You went completely into the weeds here: Authoritarians *don't* cancel
each other out, and in fact probably are more consistent and
predictable in their voting patterns than Neurotypicals.  It's also a
mistake to equate authoritarianism with more common types of
conservatism: It's a very specific definition even though
authoritarianism is a continuum and the percentage of the population
one would classify as authoritarian varies depending on where you draw
the line.

> http://en.wikipedia.org/wiki/List_of_United_States_presidential_elections_by_popular_vote_margin
> Don't forget that the political actors constantly recalibrate to insure the
> equilibrium.
> It's overwhelmingly clear that those who make uninformed choices are
> perfectly neutralized. I understand that you identify those with a tendency
> to make uninformed choices as "authoritarian" and hence there is nothing to
> improve upon.

Again, you're thrashing in the weeds here because you don't understand
the underlying concepts.  Authoritarians are frequently the *most*
informed and active voters.  It's their *biases*, especially towards
prejudice and aggression, that present problems, not their lack of
information or cognitive processing limitations.

> Only if you were an authoritarian yourself you would look at the other half
> and consider them as authoritarians.

My personal political philosophy has nothing to do with my ability to
use objective scientific measures to design systems.  In addition to
Altemeyer's book you'd also have to read mine to understand the irony
of the whole situation, and of your comment in particular.

> Interestingly you aspire (just like the others) that the "broad middle of
> the normal curve" is supposed take charge. While that is exactly what we
> have and appears to me as the problem. I have a feeling that most people are
> just out of touch with reality and don't quite understand what the "broad
> middle" really is (false-consensus bias?).

I disagree, and this "broad middle of the normal curve" is a highly
reproducible phenomenon when doing political orientation surveys.
Sure, some, maybe even large, percentage of that "moderate" population
are simply clueless.  But because their biases are weaker and less
consistent, they are actually better suited to make decisions for The
People as a whole than the relying on the people out on the tails,
which is what we mostly do now (although it's actually worse than
that, because social dominators occur on *both* tails since unlike
authoritarianism this personality characteristic is not highly
correlated to political orientation, social dominators sharing the
authoritarian tendency toward prejudice against those they consider to
be less worthy than themselves).

> I think that the future belongs to a system were decisions are made by a
> minority of very well informed people that are directly validated by the
>. I used to think that liquid democracy
> with it's powerful delegation is the best approach but I tend towards
> competition these days. The pirate party is a great place to experience the
> different dynamics.

I think we mostly agree here, although I'd put the emphasis on this
"validation" process: It needs to be the core of the system, with the
"experts" contributing concise digests of the issues that the people
on that "broad middle of the normal curve" use to make their
decisions.  They can't merely rubber-stamp a decision made by others,
they have to evaluate the arguments and the qualifications of the
"expert" contributors and then vote on the matter.  But I also see a
role for delegation and competition in this process, as you suggest.

>> What I was hoping to achieve with my posts was to get someone on this
>> list to defend "consensus" as a design goal (decision rule) and/or
>> defend the proposal to have moderators or juries who would "correct"
>> irrational or other disruptive behavior.
>
> I am drifting into a different universe than the meta gov or AG MFT people
> so I can't help you with that.

Too bad.  I think most of the MG group has finally backed off on the
consensus design goal (at least, it's not on the Home page anymore ;-)
and I was just hoping to prevent a similar waste of time exploring a
blind alley here.

> According to my understanding "consensus" or "moderators or juries" as
> central goal are inefficient and not worth pursuing.

Inefficient is the best case.  "Unworkable" or "dangerously
susceptible to corruption" is the more likely result.

> I am looking at it from much more dynamic (opt-in) perspective whereby
> consensus is implicit and doesn't need further consideration.

I'm fine with dynamic delegation as a guiding principle.  It gives us
the best aspects of our current representative democracies without
those huge flaws (monetary corruption and the fact that authoritarians
and social dominators seek out those positions).
  Regards,
    Scott

> Regards
> Martin