Entry 267 closes year eleven, and the closing fortnight supplied its own miniature of the year: Google’s Gemini launched (Dec 6) with impressive benchmarks and a hands-on demo video so fluidly responsive it briefly reset expectations, until the disclosure that it was edited (prompts abbreviated, latency cut, the interaction reconstructed), and the old doctrine (“the demo is the dream; production is the compromise”) claimed its most consequential 2023 specimen: in the eval era, a misleading demo isn’t marketing anymore, it’s a factuality regression in your own launch, caught by the planet’s review queue within 48 hours (the replication machine, now aimed at product videos). Meanwhile E3, the cathedral of exactly that demo culture, the stage of the old Sony massacre and the Keanu moment, was formally declared dead (Dec 12) after 28 years, killed by the direct-to-audience channels (Nintendo Directs, State of Play) that made the middleman theater redundant. The two obituaries are one lesson: the demo’s power migrated to the deploy (civilization-runs-the-eval, achieving total coverage). ...

The weekend resolved as the mid-crisis logic demanded: Altman restored as CEO within five days of the firing, the Shear interim lasting roughly 72 hours (his tenure’s principal artifact: a tweet clarifying the board “did NOT remove Sam over any specific disagreement on safety,” deleting the weekend’s leading theory without installing a replacement), the old board dissolved save one, a new small board seated (Bret Taylor chairing, Larry Summers arriving as the establishment’s notary), the employee letter having reached ~95% signature coverage, including, in the detail that will feed governance seminars for a decade, signatories among the board’s own allies and Ilya Sutskever, who co-signed the letter against the action he’d voted for days earlier (“I deeply regret my participation”), and Microsoft converting its weekend of leverage into a board observer seat: influence formalized, liability declined, the earlier structure’s actual power topology now documented by stress test (the file’s charter-vs-cap-table question answered: the cap table won, wearing the charter’s language). Investigations pending; the fired board’s specific cause remains unstated, which the file continues to hold as the weekend’s original sin and its enduring mystery (the eventual review’s findings, “a breakdown of trust,” reportedly, over candor in board communications, will satisfy no one, which may be the truest possible finding). ...

Filing from inside the strangest governance event this archive has ever covered, outcome unknown, per the charter (write at the moment of not-knowing): on Friday afternoon, OpenAI’s nonprofit board fired Sam Altman, four sentences of announcement, “not consistently candid” as the entire stated cause, no warning to investors including the $13B partner, no succession plan visible, Greg Brockman resigning within hours, and Mira Murati installed as interim CEO. As I file Sunday night: reporting says Altman was in the building today negotiating return terms (wearing a guest badge he photographed, “first and last time i ever wear one of these”), the board has apparently deadlocked, a second interim CEO (Twitch’s Emmett Shear) is rumored incoming tonight, Microsoft’s Nadella is reportedly working the phones with the leverage of a partner who learned of the firing one minute before the world, and 700+ of ~770 employees are preparing a letter threatening to follow Altman to a Microsoft-hosted lab unless the board resigns. The earlier pre-registration executes with terrifying precision: “governance structures reveal themselves only under governance stress, and this one will meet stress,” the capped-profit wrapper’s novel geometry (a nonprofit board, legally bound to a mission rather than shareholders, holding fire-the-CEO authority over the most commercially consequential company of the decade) has now fired its weapon and discovered the recoil: the board had the authority but not the power (the positional doctrine at maximum stakes, the employees are the company; the compute contract is the balance sheet; a mission without a workforce governs an empty building). Whatever Friday’s actual cause, and the absence of a stated one converted every observer into a conspiracy theorist by Saturday brunch (the velocity doctrine: in an information vacuum, coordination happens at group-chat speed against you), the meta-lesson is already filed: firing without a communicated cause is a resignation letter written by the board (the two-sentence-note doctrine inverted: under-communication at maximum stakes reads as either cowardice or emptiness, and both readings kill the authority that chose silence). ...

Governance fortnight, exactly as staged. Biden’s AI executive order landed October 30th, the compute threshold made law-adjacent (models trained above 10^26 FLOPs must report to Commerce: the materiel doctrine now has a number, and every lab’s training-run planning now includes a regulatory line item), plus safety-test disclosure under the Defense Production Act (the Korean-War statute as AI governance, the US regulating via supply-chain authority because Congress won’t ship, the geography confirmed), watermarking research mandates, and immigration provisions the industry actually cheered. Two days later, Bletchley: 28 nations plus the EU, including, notably, both the US and China at one table, signed a declaration acknowledging frontier-model risk, with the UK’s new AI Safety Institute announced alongside (and the US matching with its own within the week, institutional isomorphism at summit velocity). The file’s calibrated read: declarations are not regimes (the fog clause), but the infrastructure being built, safety institutes with model access, compute reporting, eval science funded as statecraft, is the part that compounds. The 2023 file will be remembered for this stack-up: capability jumps in March, governance scaffolding by November. The gap between those dates, eight months, is the new unit of institutional velocity, and it’s the fastest this archive has ever filed for anything without a body count driving it. ...

The fortnight opened with horror outside this archive’s lane: the October 7th attacks in Israel and the war now unfolding in Gaza. Per the charter this file has held for years, log what touches the craft, hold the rest with humanity, the craft-adjacent notes are grim ones: the information layer is performing worse than in any conflict this archive has covered (platform trust-and-safety teams gutted by the earlier-era layoffs meeting maximum-stakes misinformation velocity, X’s crowd-sourced Community Notes visibly outpaced by fabricated footage recycled from other wars; the information-theater doctrine now degraded by two years of moderation disinvestment, exactly as the retention thesis predicted operationally), and verification itself, the craft of knowing which video is real, is now civilian infrastructure that mostly doesn’t exist (the provenance wars arriving at their gravest use case). Hug your people; verify before sharing; the archive keeps minutes and its limits. ...

SBF’s criminal trial opened October 3rd in Manhattan, eleven months from the collapse to a federal courtroom, warp speed by white-collar standards (the QuickBooks forensics apparently self-documenting), and the opening arguments frame the exact question this archive filed at Theranos: the defense’s “math nerd who made mistakes amid chaos” versus the prosecution’s documented intent (the co-conspirators, Ellison, Wang, Singh, have all pleaded and will testify; when your entire inner circle takes cooperation deals, the adversarial-reviewer arrives as a witness list). The file’s watch-item is testimony mechanics: Caroline Ellison’s spreadsheets (the “seven alternative balance sheets” prepared for lenders, per filings) are the reconciliation doctrine as exhibit, the confession was always in the delta between the versions. Verdict expected within the month; the archive pre-registers guilty on the core counts with the confidence of someone who read the first-day filing and holds, per its own doctrine, that the interesting question was never the verdict but whether the industry metabolizes the lesson (custody, governance, boards) or just the villain. History suggests the villain (Enron produced SOX and two decades of “this time is different”; the file expects both again). ...

Unity, the engine under roughly half the world’s games, announced a runtime fee September 12th: per-install charges, applied retroactively to already-shipped games, calculated by Unity’s own opaque telemetry, hitting hardest the exact indie-and-mobile studios whose success stories built the platform’s brand. The developer revolt was total and instant: studios publicly pledging engine migration (Godot’s donations spiked 4x in a week), collective open letters, some devs disabling Unity ads mid-monetization in protest, and, the detail that graduated it from pricing dispute to trust collapse, the company had quietly deleted its GitHub repo tracking terms-of-service changes months earlier, the receipts-vanishing move that converts customers into archivists (the emails-are-forever doctrine: the internet had the diffs anyway). The file’s structural read, and it’s the Reddit lesson wearing an engine license: platform pricing power is real but bounded by migration cost, and retroactivity is the one move that revalues migration cost overnight, a fee on the future is a negotiation; a fee on the past is an expropriation, and developers reprice platform risk (not platform price) accordingly, permanently (the graveyard’s common headstone reads “the terms changed underneath us”). Walkback is already in motion (apology posted, revisions promised, the CEO’s exit within the month is this file’s confident pre-registration); the trust, per every precedent in this archive, follows a different and slower curve than the pricing. ...

A working fortnight (the news cycle’s kindest gift in a while), spent shipping the thing I want on the record at length because the file believes it’s where the whole industry lands within two years: eval-driven development for AI-touched systems. The earlier regression-catch converted our leadership; the quarter’s mandate followed; and the pattern we’ve converged on deserves its Proverbs entry: every AI-assisted feature ships with (1) a golden set, curated input/expected-judgment pairs, owned like tests, reviewed like code; (2) an LLM-as-judge layer for scale, calibrated against human ratings quarterly (the judge drifts too, the context-psychology applies to evaluators; who watches the watchmen: a rubric, versioned); (3) regression gates in CI, a model update, prompt change, or retrieval tweak that moves the golden-set scores blocks the deploy exactly like a failing test; and (4) production sampling with human review, because the golden set is the map, not the territory (the funnel-leaks doctrine: production is where the distribution actually lives). None of this is novel research; all of it is novel discipline, and the gap between teams that have it and teams demoing vibes is about to become the gap between rehearsed orgs and everyone else’s incident reports. The 2013 kid tested code; the 2023 principal tests judgment, the asset was always judgment, now with CI gates. ...

Spain won the Women’s World Cup yesterday, a golden-generation coronation, La Roja’s first star, won 1-0 over England behind a midfield (Bonmatí conducting) that plays the sport’s most beautiful software, and within minutes the story became the federation president kissing a player on the mouth during the medal ceremony, without consent per her immediate testimony, followed by a non-apology the file grades as the first draft of worse to come. The enforcement-consistency file and the values-pricing file converge at maximum visibility: the institutional-power script (deny, minimize) is already executing in public against a workforce that just delivered the institution’s greatest achievement. The file’s read, calibrated by a decade of these entries: the trophy gave the players leverage no negotiation could (the positional doctrine at its purest — the product withholding itself remains the only veto that always works), and so the file pre-registers the sequence: mass selection-refusal within the fortnight, formal proceedings shortly after, and the resignation, eventually, entirely on the players’ terms. The sport’s best story and its oldest one, in a single ceremony. ...

LK-99 has delivered the best public science theater of the decade, whatever the verdict: two weeks of global, real-time, open replication, labs on four continents racing syntheses, Twitter threads with X-ray diffraction data, simulation teams publishing band-structure analyses within days, a Russian anime-avatar account growing crystals on a kitchen bench, and betting markets repricing hourly (peak: ~60%; as I file: ~8%). The emerging consensus, copper-substitution artifacts and a cusp-like diamagnetic response, with Cu₂S impurities explaining the resistivity drops, is arriving through exactly the process this archive has spent a decade celebrating (adversarial review, blind-injection culture, public falsifiability): extraordinary claim, total transparency, distributed verification, fast honest death. Compare the timeline against cold fusion’s years of murk (1989) and the file’s conclusion writes itself: open science with preprint velocity and social-media coordination is the best error-correction machine humanity has ever run, it just looks like chaos while it runs (the March-Madness variance, but for condensed-matter physics). The floating rock was not the future; the replication machine is. File closed with gratitude. ...