The most clarifying sentence written about the maxxing internet this month came from someone who wasn’t writing about the maxxing internet at all. Karina Nguyen — an AI researcher whose product work at the frontier labs gives the observation unusual weight — posted a thesis about where consumer AI goes next, and put looksmaxxing in the middle of it:
the 2nd wave of consumer ai is priesthood. i think people just want an oracle that tells them who they are, what to believe, how to look, who to love, if they’re healthy, where they belong. that’s why astrology, religion, looksmaxxing, wellness, identity will be the greatest consumer ai obsessions. it’s less rational, far more intimate, and will matter more as the knowledge work disappears and stops being the primary identity/status signal. meaning is a larger market than code
— Karina (@karinanguyen) July 24, 2026
Twenty-one thousand views — a rounding error next to the numbers we usually track. We’re covering it anyway, because it’s the rare post that explains a month of news better than the news explained itself.
The scene built the prototype
Run this month’s stories back through the oracle frame and watch them click into place. Clavicular’s LooksMax AI app — scan your face, your meals, your text messages, receive judgment — isn’t a gadget; it’s a confessional booth with a subscription tier, and it allegedly charted top 10 in a day. Omoggle is the oracle as gladiator pit: two supplicants, one verdict. The PSL scale was the hand-computed, pre-industrial version — a priesthood of forum posters issuing numerical fates. The looksmaxxing internet has been running oracle infrastructure for a decade. The AI industry is only now noticing that the demand it served — tell me who I am, tell me my number — generalizes to everything.
And notice which of Nguyen’s categories got to production first. Not astrology, not wellness: “how to look.” The face was the beachhead, exactly as the clinicians warned — because it’s the question young men were already asking strangers on the internet, and the one they’ll pay an app to answer privately.
Meaning versus code
The timing gives the thesis its bite. This same weekend, the Journal declared corporate tokenmaxxing over — enterprises are done overspending on AI to write code and summarize meetings. Set the two posts side by side and you get a complete map of the moment: the rational market for intelligence is entering its thrift era, while the irrational one — identity, belonging, the mirror — is just igniting. “Meaning is a larger market than code” is either a profound product insight or the grimmest sentence of the year, and nothing prevents it from being both.
It also names the stakes better than we usually manage. An oracle that tells a 16-year-old what his face is worth isn’t a neutral product category — it’s the harm the researchers documented, operationalized at consumer-app margins, with the biometric questions attached. Delphi at least made you climb a mountain first.
Where this lands
The original oracle at Delphi ran for a thousand years on one product: certainty, dispensed by authority, about questions with no checkable answers. It had a priesthood, a fee schedule, and a captive audience of the anxious powerful. The ruins are lovely this time of year.
The maxxing internet rebuilt the model from scratch without reading the classics — forums as priesthood, PSL as scripture, and now apps as the temple. If Nguyen is right, the rest of consumer AI is about to follow the path the looksmaxxers walked first. The scene we cover wasn’t a subculture after all. It was a pilot program.
