Complexity made slightly less complicated
July 2, 2019 12:41 PM   Subscribe

Complexity explained offers a set of articles, systems, and interactive toys that explain aspects of the new field of complexity science, from the emergence of the blob, to the reasons for traffic jams, to the dynamics of herd immunity.
posted by blahblahblah (24 comments total) 51 users marked this as a favorite
 


I was wondering the other day if you put a display of the current odometer reading on the back of everyone's car would that help everyone keep an even flow going.
posted by bleep at 1:17 PM on July 2, 2019 [1 favorite]


I'm confused, bleep - how would that help?
posted by Greg_Ace at 1:27 PM on July 2, 2019


I was wondering the other day if you put a display of the current odometer reading on the back of everyone's car would that help everyone keep an even flow going.

I imagine autonomous vehicles could be programmed to modulate their speeds in an automatic fashion, so as to minimize their contribution to traffic jams coming up the road.
posted by They sucked his brains out! at 1:49 PM on July 2, 2019 [3 favorites]


If I see everyone around me is going around 30 I'll try to keep it around 30 too. Otherwise I'm just guessing about how fast I should be going, hence the jam.
posted by bleep at 1:52 PM on July 2, 2019 [2 favorites]


I'm confused, bleep - how would that help?

Probably that was supposed to be speedometer, not odometer.
posted by mhum at 1:55 PM on July 2, 2019 [3 favorites]


Got it, that makes more sense to me. Thanks.
posted by Greg_Ace at 2:00 PM on July 2, 2019


My googlefoo fails but there was a blog post by an engineer that found just one or two cars leaving a buffer and not speeding up slamming the breaks calmed large sections of rush hour, I think it was on the 405 around Seattle. One theory is that a small threshold of SDC autonomous cars that maintain buffers will improve existing traffic situations.
posted by sammyo at 2:24 PM on July 2, 2019 [2 favorites]


You have two choices of networks, the Erdős–Rényi (ER) and Barabási–Albert (BA) network.

Do either of these choices explain why all mathematicians are Hungarian, or at least Hungarian-adjacent? I mean, my Dad had Opinions on this, but it wasn't like he systematised them or anything.
posted by Joe in Australia at 3:43 PM on July 2, 2019


sammyyo, was it this jam busting electrical engineer?

I loved his video and whole traffic experiment website and spent many a traffic jam in Seattle exploring his theory. I didn’t seem to ever bust any traffic jam waves but loved his advice that it really didn’t matter how many people you let in front of you. Just keep a large buffer in front of you at all times so you can go a steady speed.

For a time I lived in Ballard and my girlfriend lived outside Bremerton so we were in traffic all the time trying to reach the other one. We would play around with his challenge to never hit the brakes in a traffic jam and go a steady speed. If nothing else it was entertaining but it really pissed people off when you don’t tailgate a car when there’s a traffic jam. This annoyance at the irrational behavior of Seattleites overrode any novelty at trying to bust a traffic jam.
posted by the thorn bushes have roses at 4:25 PM on July 2, 2019 [3 favorites]


Thanks, this is great!

Complexity fun to think about. 2 + 2 = 4, right? Is there a minimum of complexity required for a system to solve that? You need an apple pie universe, with someone to ask the question, who has concept of a number system, and an implementation of that concept. Given that the complexity of the solution depends on the concept and implementation, is there an absolute minimum complexity for a system solving 2+2 that is actually built into the universe?

There are surely others who have thought about this more than I, but it keeps me off the streets.
posted by cowcowgrasstree at 4:32 PM on July 2, 2019 [1 favorite]


It’s not so much that all mathematicians are Hungarian-adjacent, it’s that they’re all Erdős-adjacent
posted by TheShadowKnows at 5:07 PM on July 2, 2019 [7 favorites]


Do either of these choices explain why all mathematicians are Hungarian, or at least Hungarian-adjacent?

I'm not saying it was aliens, but Leó Szilárd said it was aliens.
posted by clew at 5:26 PM on July 2, 2019 [6 favorites]


sammyyo, was it this jam busting electrical engineer?

Oh, man, I was just thinking of that same writeup when I clicked in! I find myself doing little bits of that in middling traffic sometimes, though I rarely feel up to committing to it entirely because, as you say, people don't always seem interested in participating in non-normative driving experiments.
posted by cortex at 5:45 PM on July 2, 2019 [1 favorite]


I feel one thing about vaccines that never gets talked about is the fact that they often aren't very effective on an individual basis. The BCG vaccine against Tuberculosis for example, varies widely in effectiveness - some studies show only 20% of vaccinated children were actually granted protection against being infected. But the boosted immune system also halves the chance that the disease progressive to an active state, at which point it can infect others.

Doctors in particular need to get vaccinated against certain diseases, and because it's important enough, they get tested for antibodies after the fact and it's common for the vaccine to be ineffective, at which point they can get re-vaccinated to try again, and sometime after enough failures they just conclude that this doctor will never become immune because for whatever reason the vaccine just doesn't work.

Basically vaccines aren't some super armor where you take it once and you're immune. Even with a 100% vaccination rate, the actual immunity rate is much lower and like the models above we rely on partial herd immunity to a large degree to minimize and contain the infection.

I've had the BCG vaccine and I still wouldn't want to hang around people with active tuberculosis because chances are I'm not actually immune and with repeated exposure I'd probably get infected. We had prolonged contact with a person once, but theirs infected the lymph nodes rather than their lungs.
posted by xdvesper at 6:21 PM on July 2, 2019 [1 favorite]


One theory is that a small threshold of SDC autonomous cars that maintain buffers will improve existing traffic situations.

It's easy enough to simulate: the idea is counterintuitive, perhaps, but aims at giving drivers ahead of you time and space to increase their average velocity, which in turn allows you and those behind you to do the same.
posted by They sucked his brains out! at 8:42 PM on July 2, 2019 [1 favorite]


That explanation of a traffic jam is a total LOL - it doesn't explain a traffic jam at all. At best it explains that slowing everyone down (or more likely getting everyone to drive a consistent speed with consistent acceleration) would lessen theoretical traffic jams, but traffic is not theoretical like a stupid race- it represents people getting to actual destinations, some of which have competing priorities (like for example ambulances and late moms and ages (little old ladies)). I really wish traffic engineers would stop it with explanations like that - think about things in stupid ways and come up with stupid solutions. In short, who cares about average speed except for idiot traffic engineers playing with stupid models.
posted by The_Vegetables at 8:37 AM on July 3, 2019


An implementation of addition doesn't require a concept of number. Rocks rolling down a slope will implement addition just fine without positing that somehow the mountain knows about numbers. The concepts are posthoc abstractions that we come up with to explain that rocks can be found in different configurations, and many (all?) of them will break down if scrutinized too closely.
posted by I-Write-Essays at 10:00 AM on July 3, 2019


the new field of complexity science

This seems like a surprising claim to me. As I recall, the big first fad was in the mid-90's, eg the popular science books by Holland or Gell-Mann, the Santa Fe Institute, etc -- heck, Horgan's 1996 book The End of Science had a (somewhat but not entirely misguided) chapter on "The End of Chaoplexity". It's hard to tell from that overly-complex web page what's happened of significance in that world since the first boom 20 years ago.
posted by chortly at 11:35 AM on July 3, 2019


seconding chortly, Prigogine's relevant work was as early as the 1950s, his Nobel in the 1970s, and his popular books started in the 1980s.

OTOH, from the point of view of the history of science, the 1950s is new-ish and widely available cheap simulations are yesterday, and there's plenty of new work to think about, so no more than a slight eyeroll. Less than that; all my eyeroll is for the implication that we should pay attention to something because it's new.
posted by clew at 12:01 PM on July 3, 2019


Ha! Prigogine, Kinetic theory of vehicular traffic, 1971. Shoulders of giants are just visible over the headrest in front of you.
posted by clew at 12:03 PM on July 3, 2019


> That explanation of a traffic jam is a total LOL - it doesn't explain a traffic jam at all. At best it explains that slowing everyone down (or more likely getting everyone to drive a consistent speed with consistent acceleration) would lessen theoretical traffic jams, but traffic is not theoretical like a stupid race- it represents people getting to actual destinations, some of which have competing priorities (like for example ambulances and late moms and ages (little old ladies)). I really wish traffic engineers would stop it with explanations like that - think about things in stupid ways and come up with stupid solutions. In short, who cares about average speed except for idiot traffic engineers playing with stupid models.

Similarly I find myself bemused by claims that improving throughput makes traffic worse, and getting rid of capacity improves traffic. I mean, *sure*, but people are driving for a reason right? I haven't seen it demonstrated whether the imposition of an untenable commute, for example, results in people finding jobs closer to home that pay as well, etc. or just hurts poor / working class people.
posted by mikek at 4:10 PM on July 3, 2019


I didn't mean to sound dismissive in my previous comment -- I was last reading around in this field 25 years ago, and would in fact be interested to hear what new fundamental discoveries (as opposed to just new applications of existing models to new domains) have happened since then, if anyone knows. As I recall, the dream was always a "science" of complexity, which meant not just accurate computer simulations or simple rules that give rise to complex dynamics, but rather a broader over-arching theory of complexity that unified all the disparate results in the way Newton or Einstein's theories did. Adding preferential attachment to the standard random network model, eg -- which does seem to have happened post-1995 -- doesn't really seem like the sort of deep unifying progress they were hoping for.
posted by chortly at 5:36 PM on July 3, 2019


Similarly I find myself bemused by claims that improving throughput makes traffic worse, and getting rid of capacity improves traffic.

By one account, the Katy freeway in Texas is the world's widest, and Houston's traffic is as bad as ever. Adding capacity does not appear to reduce traffic.
posted by They sucked his brains out! at 8:52 PM on July 3, 2019


« Older “It’s at odds with the language we use to talk...   |   Girls Rush Out Like Ghosts Newer »


This thread has been archived and is closed to new comments