The Radical Technology of Christopher Alexander

Nice piece on Christopher Alexander.

Seb Paquet found a good critique of the documentary I was so disturbed by.

Brilliant!

I HATED episode 2 of All Watched Over By Machines of Loving Grace

We just watched it in class today and I ended up screaming at my colleagues. I was furious with it and with them for following it. I think I even used the word "dishonest" which got me an arch look from the tutor.

Now I've just got off an hour on the phone bitching about it with my friend who's another systems theorist / cybernetics fanboi.

Anyway, something that got under my skin that badly must be doing something right. Yes, there are questions to be asked about what assumptions have historically gone into systems theory. There are definitely issues with how it's been applied and put to work politically. Yes, there are things to think about when you deny that power exists (Tyranny of Structurelessness etc.)

I still hate it. But I think it's good for me to be forced to work through a response.

Kill Math seems a project worth watching.

Update : this is excellent. (Though there might be an issue when you get to higher than two dimensions.)

Over on Quora I'm considering the hypothetical libertarian traffic planner.

I like to consider the case of a hypothetical "libertarian road traffic planner" (assuming a libertarian could be persuaded to take up such a job, but let's assume that she's working for a private city and the money's too good to resist)

The planner is faced with the problem that every morning and evening there's a major traffic-jam which means that the average drive-time to and from work is about two hours, when it should really be about 40 minutes. How can she reduce it?

Because the libertarian doesn't believe that phenomena have collective (or emergent) causes, she assumes the problem must stem from individual failure. She notices that certain drivers, who are perhaps a bit more skilful at driving, more aggressive in challenging other cars at junctions, are more decisive and less risk-averse under pressure, more willing to drive fast or cut corners etc, are able to beat the average and get home in about an hour and a half.

She therefore thinks she could cut 25% off the average drive-time if *everyone* could be persuaded to improve their driving. How to do that though? Obviously, people need to take more driving lessons and practice harder. But as a libertarian, she doesn't want to force people to do anything, so she'd better provide incentives to encourage them. Better yet, incentives in the form of removing unfair restrictions.

The obvious thing to do, therefore, is to eliminate speed restrictions. Allow everyone to drive as fast as they like. That means that the real experts won't be held back, there'll be a greater reward for their investment in their skill. And that, in turn, will create a greater incentive for everyone else to learn better driving skills too.

That, then, is our well-intentioned libertarian's response to road congestion.

Now suppose we need to argue against the libertarian? How would we do it?

a) We'd point out that traffic congestion is not a simplistic scaling up of individual failure. There are emergent, non-linear, turbulent effects when a lot of people try to access the same resources.

b) We'd point out that some of the interactions in driving, such as the challenges for priority at road junctions, are zero-sum games. Hence, one driver can't win the junction (and get home quicker) without the other driver losing it (and NOT getting home quicker)

c) we'd point out that not everyone can ever aspire to being as good as the fastest drivers. The elderly, those with certain physical disabilities. Those driving children who they love and want to protect will remain more risk averse.

d) we'd point out that the increased number of accidents caused by the increasingly risky behaviour of the "elite" will block roads and slow everyone's journey down.

e) we'd point out that this solution misses many other options that could improve the transport situation in the city far more dramatically (everything from, on the one hand, building more roads, to, on the other, providing more buses).

f) In short, we'd point out that IT WON'T WORK to reduce travel time except for an infinitesimal minority of super-drivers and will cause more trouble for everyone else.

Of course, the libertarian might simply be too ideologically fixated to accept any of these arguments. She may not accept that there are non linear effects in many-car-interactions. She may have heard that in real life there are no such things as zero sum games. She may have read some garbled account about somewhere in Holland where they took away all the road signs and people drove safer. She may think that any top-down scheme (such as road building or bus-providing) must of necessity be less efficient than her bottom-up scheme.

She may, in the last resort, fall back on saying that ultimately, average speed doesn't matter. The most important principle is to remove the restrictions unfairly holding back the best drivers. (Although this is weird in my contrived example as that's her job.)

So, basically any argument against the libertarian bifurcates on one of two trajectories.

1) A libertarian who doesn't care about the welfare of society as a whole, just the freedoms of those powerful enough to enjoy themselves, regardless of the consequences for everyone else.

This kind of libertarian is just special pleading for a particular interest group and there's no reason to take her more seriously than someone who claims to be the true heir to the Tsar of Russia and wants your help getting their empire back.

What's in it for you?

2) A second kind of libertarian who insists that the freedom she wants is going to benefit society as a whole (or at least, the majority of it).

In this case you can get down to details about *how* everyone is expected to benefit.

Does their argument make naive assumptions about how individual self-improvements scale up to general social welfare?

Does it assume that everyone can get the benefit of things that are actually zero-sum competitions for scarce resources. (Eg. "wealth" in the most vague and hand-wavey sense isn't scarce, but concrete opportunities for wealth such as "money" or "market share at this moment" or "jobs in this town during my lifetime" etc. often are)

Does it ignore or dismiss opportunities for "collective" solutions? Or if it accepts that some benefits are possible from collective actions, does it plausibly demarcate good from bad?

Etc. etc .

Ben Hyde:

I learned this rule from “Reading the Forested Landscape,” if you stand in the forest and all the trees are about the same size it tells you when the catastrophe happened. That works just as well for housing developments. If you take the rule seriously then one of your first questions is always: what were the catastrophes?

++ThoughtStorms

System Thinking