Three Great Pieces
People watching, DHCP protocol fiction, the election-losing theory of democracy
My jetlag has segued into a bad cold, so I’m tabling the post I had planned. I did read three very good things in the last week that I want to share though.
First, this post by
on people-watching, which might be the best bit of non-fiction I’ve read so far this year.
A lot of people try to write listicles like this, but they’re usually no more than lazy inventories of unexamined personal prejudices, bolstered by a decided lack of actual observational talent. This is different. The insights ring both grounded and surprising, in the sense that they seem genuinely born of a preternatural skill at people-watching, and take seemingly obvious points just a little deeper than normal people can. Objective evidence, if you want it, can be found in Shani’s drawings and sketches, casual outtakes from her professional wedding painting work. Proof-of-Eye if you will.
I think I’m a pretty good people-watcher myself, and in fact much of my consulting work rests on that skill (though mine rests more on reflective analysis than acuity of observation), but this is next level. I found myself thinking both, I want to learn from her and, I don’t want to ever be observed by her.
Next up, the third place winner from the protocol science fiction contest we just ran, by
.
The challenge of the contest was to spin a strange story around a familiar technological term related to protocols. DHCP is the protocol your devices use to connect to the internet via your ISP. This story really delivers, cashing out the logic of the design pattern underlying DHCP as a weird tale of political rebellion and address-space scarcity/inequality. This is exactly the sort of story I was hoping for when we framed the contest. You can feel your protocol literacy going up as you read, and your appreciation for the politics of architectural patterns leveling up.
And finally, a brilliant classic account of the nature of democracy, the first chapter of a 1991 book by political scientist Adam Przeworski, Democracy and the Market.
The chapter is here (paywalled; find yourself a PDF by your preferred techniques).
The book itself is not of any particular interest in 2025, but this first chapter was apparently very influential in political science. The reputation is well-deserved; it certainly subtly subverted my mental models of what democracy is. The account rests on a deceptively simple-seeming definition of democracy as a system in which parties lose elections. Building on this, it then uses a clever game-theoretic analysis to ask and answer why losing parties have an incentive to accept unfavorable outcomes.
The model explains to my satisfaction why MAGA didn’t accept the loss in 2020, which therefore serves as a successful prediction of the 1991 theory.
Tldr: Acceptance rests on feeling like you’re represented robustly enough in the infinite game that you’ll have future shots at favorable outcomes, and that disrupting the system in response to a single loss is not worth it in NPV terms. In 2020 at least, I think MAGA still felt like they were part of a one-time rebellion that had to hold on to power or be eliminated from the infinite game forever. I think they still feel that way despite winning another election — there is an apocalyptic, “once and for all” tenor to their frantically inept solutionism. Their embrace of monarchism is less a rejection of democracy than a sense of the unviability and unreasonableness of their own political desires within its rules of engagement.
The Przeworski chapter, a new favorite, pairs very well with an old favorite you’ve probably seen me cite before: Frank Chimero’s Only Openings, which contrasts an apocalyptic-endgame “wolf slaughter” approach to complex problems and a sensitive “bear management” approach (the terms refer to two historical modes of wildlife management at Yellowstone National Park).
MAGA, and the late-not-lamented DOGE are both wolf-slaughter philosophies in Chimero’s terms. They treat experts, bureaucrats, and institutionalized knowledge as existential threats and pure deadweight liabilities, to be rooted out and destroyed utterly. Musk in particular sounded Thanos-grade deranged about it early in his short DOGE chapter declaring “this is our last chance to root out bureaucracy” or something. The attitude extends to the rest of Silicon Valley, with its PMC-derangement syndrome (the big brother of woke/DEI derangement syndrome; PMC stands for James Burnham’s “Professional Managerial Class” Ayn Randian boogeyman).
A sense of existential doom can trigger an utterly panicked instinct to wholesale slaughter of a sort that can devastate entire human ecologies beyond repair. MAGA may have succeeded with this program at the Federal government already, at least to the extent the 19th century wolf-slaughterers of Yellowstone did.
But all is not lost. The gray wolf has made a recovery in Yellowstone, and a thriving ecology is beginning to get established again. The PMC — an apex predator species in governance ecologies at least as valuable as billionaires —may yet return from the brink of extinction. I hope it doesn’t take a century plus though. I’ve come to see my mission running the Summer of Protocols as being in part about restoring the legitimacy of procedural stewardship as a societal calling, and updating institutional and individual archetypes associated with industrial bureaucracies for an age of networked protocols.
Much as I disagree with and detest the MAGA/DOGE agenda, I can see where it is coming from. This is two sets of people who can literally see no future for themselves unless certain other people are eliminated “once and for all.” Henry Jenkins’ Civic Imagination project at USC found that MAGA types were so hopeless about gains made by various minorities, apparently at their expense, they couldn’t imagine (say) their own condition in 2050. Something similar is true of the sincerely (rather than mercenary) Trump-supporting tech oligarch class. They think the “PMC” will exterminate them unless they do it to them first.
Something similar is true of the far left — they can literally see no future for themselves unless billionaires and “racists” are wiped out “once and for all.”
These are all PTSD responses to threats to overidentification affiliations, as I argued to Jenkins when we chatted about his research in 2019. Which is an explanation but not a justification. The wolf slaughterers still need to be stopped. Bear managemebt literacy still needs to be installed in the civic imagination.
Once and for all. It’s a phrase I’ve come to really detest. It invariably points to slaughtering instincts and unforced eschatological endgames that devastate shared ecologies. The particular flavor doesn’t even matter. As many have noted, Trumpism increasingly resembles Maoism rather than Hitlerism. It also now practices a more cartoonish variety of anti-meritocratic right-wing DEI than the left ever did. Does it matter? Within once-and-for-all ideological design spaces, these are distinctions without differences.
I found the Przeweski book chapter via this post by
which is also worth reading. It makes an interesting point about Brian Eno’s music being about rules that open up rather than constrain possibility.
This is something the tech world sadly seems unable to see at the moment. Blinded by PMC derangement syndrome, enthralled by deeply silly theories of “elite human capital,” and still unchastened despite the disastrous tariffs rollout, they continue down the road of fundamentally juvenile political theorizing, oblivious of the hazards. And blind to perhaps the most fascinating property of complex systems — adding rules can actually increase generativity, while removing them can paradoxically destroy it.
Why do seemingly smart people who’ve built some of the most complex systems the planet has ever seen make such deeply silly analytical errors?
I believe I have an answer: They’ve mistaken the bounty of 70 years of Moore’s Law dividends and the unreasonable effectiveness of “simple” network effects driving a period of “exponential” progress for evidence of the evils of designed complexity. This is a bit like witnessing a nuclear explosion and concluding sustained power-generating nuclear reactors are evil and against nature.
Alloyed with deep conviction in their own “elite” superiority and illiteracy in reading what the “PMC” actually does, this belief structure leads them to hack and demolish working systems until they get to meltdowns and explosions all over the place.
But I suppose the only way out is through.
I found myself agreeing with everything in part 3, but still wonder how relevant it is today. The analysis seems steeped in analysis of social systems over the past ~300 years — that is, it is historically well informed. However, what if contemporary tech changes many of the rules within our set of complex systems? It may not invalidate the analysis, but it may render it irrelevant.
If aliens and God maerially presented themselves to the populace in 2025, does political theory rooted in priors even matter?
Amazing article. What do you think about the idea that these kinds of once-and-for-all terminal goals are actually instrumental goals in some historicist meta-game? I think some people in the coalition you’ve described above believe in some form of this.