[FoRK] Do You Belong to The Church?

James Tauber jtauber at jtauber.com
Thu May 6 09:00:56 PDT 2010


On May 6, 2010, at 11:32 AM, Jeff Bone wrote:
> Okay, this is tickling some interesting geometrical speculation.  Not quite got my head around it yet, but the gist would be something like:  really, to evaluate politics, we need *two separate spaces / coordinate systems.*  One representing "cause" (need better word, really preconditions or something like that) and the other "effect" (or postconditions, etc.)  They are respectively n- and m- dimensional, and there's no constraint that n and m have the same dimensionality.  There's some non-trivial (set of?) projection(s) from one into the other that have various lossy or lossless qualities depending on the translation and the mapping functions.
> 
> Of course this might be interesting and informative but would be utterly useless in general since most of the public still thinks "left / right" is a useful taxonomy and a large number are infected with the utterly ludicrous mytheme that the far "left" and "right" are identical, i.e. the space is cylindrical / wraps.  I.e., they live only in your projection target space, and don't understand that the source space these things sit at extrema of orthogonal axes.

As I've built a lot of binary games the last few years (Cats or Dogs, Potter Predictions, Typewar, Lost Predictions, Apple Predictions) I've actually been thinking about putting together my own "political compass"-style quiz with a view to doing some analysis. FoRK would likely be a nice place to gather both questions to ask and also ideas for analysis and application of machine learning.

I might see what I can whip together this weekend. So as a preliminary to that: any suggestion as to whether this should be done as (a) given statement user picks the usual strongly agree thru strongly disagree choices; (b) given two statements the user picks which they agree with more (if at all); (c) some other approach?

>> Well, I'd combine the two and say it's the scrappy bootstrap entrepreneur that hires freelancers rather than employees. 
> 
> My own assertion from last night has been gnawing at me some.  It --- along with your own extension --- actually suffers from some myopia induced by the industry we work in.  It's difficult to see how certain kinds of business --- say, the construction and ongoing operation of data centers and / or large-scale applications that use them, or perhaps the manufacture of various complex finished goods (or even parts) such as chips, computers, cars, and so on can be accomplished without some form of business that inevitably resembles "big" business.  The inefficiencies, coordination overhead and capex / opex requirements and characteristics of such businesses would seem to preclude a totally-dispersed, loosely-coupled, bootstrapped (i.e., no external capital or financing) enterprise for these types of businesses.

Isn't that at the heart of Coase's work on the nature of the firm? (incidentally, are the same reasons applicable to levels of government too? that is, why we can't just have local government?) I still have both Ostrom and Williamson on my reading list.

I'm not suggesting "scrappy bootstrap entrepreneurs hiring freelancers" is always possible, just that it's the closest to ideal free markets (and I say this in particular to contrast with those that think big businesses and investment bankers are the pinnacle of free markets). Divergence from this ideal is often required.

I agree that we perhaps suffer from a certain myopia in that this ideal is far closer to being achievable in our own industry than others. But I think the myopia also exists on the flip side when people draw too close a parallel between modern knowledge workers and 19th factory workers.

> Ricardo Semler's probably got the best take on how to do large-scale business of the industrial kind without it having the usual qualities of big business, at least from the perspective of labor.  Perhaps he should be regarded as the prototypical new-style industrial capitalist.
> 
> (Has anybody read any of his books?  I've been a fan for over 20 years, since the first printing of his first book)

Yeah, I'm a fan but share your questions on how it works at the small scale I'm operating at.

James


More information about the FoRK mailing list