Playbook
Tasteful Skills
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.
Back to Podcast Digest
Three Model Organisms For Taste - By Scott Alexander
TL;DR
Flags are Scott Alexander’s ‘model organism’ for taste — he uses online vexillology fights to show how rules like simplicity, no text, and the rule of tincture can harden from historical constraints into fake universals.
The anti-Reddit-vexillology case is brutally concrete — medieval battle flags like Venice’s insanely ornate banner broke today’s ‘good flag’ rules, while modern examples like California, Brazil, the Vatican, Iran, and the UN also violate the canon and still work.
His real target is minimalist taste posing as timeless truth — Scott argues the past 200 years pushed prose, architecture, dress, and now flags toward plainness, and asks whether people are defending soldiers and Betsy Ross or just channeling the zeitgeist.
Plot-hole culture exposes another kind of taste hierarchy — he contrasts obvious absurdities like hiding Luke Skywalker with Vader’s stepbrother on Tatooine without changing the surname against nerdy continuity checks like Ultraman’s 1,000-foot laser firing from above the 1,250-foot Empire State Building.
Tech names reveal a taste for resisting ‘easy wins’ — he says ‘Infinita’ feels like a lazy attempt to sound cool by stapling an ‘a’ onto ‘infinity,’ while ‘Vitalia’ works because it also points to biotech, longevity, and Vitalik Buterin.
His strongest emotional example is bad AI poetry — the problem with lines like ‘The sun will rise again and so shall you’ isn’t just awkward wording but the sense that the poem is manipulating the reader with stacked cliches that are too cheap and too effective.
The Breakdown
Flags as the cleanest lab experiment for taste
Scott opens by arguing that flags are a perfect “model organism” for aesthetics: simple enough that ordinary people can have opinions, but rich enough to surface big questions about history, context, and taste. He lays out the familiar Reddit-vexillology commandments — simple enough for a child to draw, no text, no busy imagery, and never “metal on metal” or “color on color.”
The internet war between Indonesia and Venice
He sketches the now-familiar online ritual: a US state redesigns its flag, and internet flag nerds pounce on every violation, praising minimalist saints like Indonesia’s red-over-white bicolor. Then he brings in the backlash — memes, mockery, and the “resistance” defending maximalist flags like California, Maryland, Brazil, Iowa, the Gadsden flag, Tibet, and especially medieval Venice with its lion, saints, ships, coats of arms, and text-packed chaos.
Why the standard flag rules may just be historical leftovers
The pro-rule camp says flags had to work through smoke and dust, be sewable by “Betsy Ross,” and drawable by patriotic children; the anti-rule camp says none of that really survives contact with history. Scott’s punchline is that medieval flags themselves were often far more ornate, modern printing kills the seamstress objection, and kids would absolutely rather draw a winged lion than “a red square on top of a white square.”
Minimalism sneaking in through the back door
His broader objection is that the whole conversation may be less about battlefields than about modernity’s drift toward minimalism. Prose got less flowery, architecture less ornamented, clothing plainer — so of course flags are next — and he asks, with some bite, whether people are really defending Betsy Ross or just being “seized by the zeitgeist and dragged off.” He caps the section with an absurdly ornate Venetian-style US flag proposal by Exuper Sven Etienne as the logical joke-solution.
Plot holes, nerd scrutiny, and what ‘paying attention’ even means
He then pivots to movie plot holes, brushing past the overdiscussed “why didn’t the eagles fly Frodo to Mordor?” and landing on a cleaner example: Obi-Wan hiding Luke with Vader’s stepbrother in Vader’s hometown and keeping the name Skywalker. It’s obviously dumb, he says, but it doesn’t diminish Star Wars for hundreds of millions of viewers — which raises the question of whether taste means rewarding hidden coherence or just becoming an “Uber nerd” who tracks Ultraman’s blaster range across issue numbers.
Tech company names and the sin of trying too hard
From there he jumps to startup naming, using Infinita as a small but revealing offense. “Infinity” is cool, he says, and that’s exactly why “Infinita” feels cheap — too easy, too obviously fishing for awe — whereas Vitalia works because it also gestures toward vitality, biotech, longevity, and Vitalik Buterin. The key word he keeps circling is “manipulating.”
Bad AI poetry as cliché-engineered taste failure
He ends on an AI poetry Turing test and quotes lines he finds ghastly, including “The sun will rise again and so shall you.” The issue isn’t merely a few weird word choices like “amassed”; it’s that the poem feels built out of prefab emotional triggers, “stringing too many clichés together” in a way that slightly works and therefore feels even more offensive.
Share
Keep Reading
Two Ways to Follow
The Weekly Echo. The inbox-shaped summary of what mattered.
New editorials announced here.
