Section
Model releases, benchmarks and market caps — dispatches from the universe where the underdog took the crown. The market question, its odds and its resolution date are real; the reporting is AI-generated satire.
1.1%Google ended July with the best AI model on Earth, a 1.1% shot — confirmed by the Bureau of Vibes and Evals, citing 'a very good demo' and rising numbers.
July 17, 2026 · 1-in-91 — Ivo Sandhu · Tech Desk
1.6%A benchmark upset nobody priced in: the end-of-July model race went to a challenger, and the leaderboard rewrote itself overnight.
July 20, 2026 · 1-in-65 — Ivo Sandhu · Tech Desk
0.8%A 0.8% longshot lands: the Bureau of Vibes and Evals ranks one model above every rival at the July 31 bell, and nobody quite believes the scoreboard.
July 21, 2026 · 1-in-125 — Ivo Sandhu · Tech Desk
0.8%A single-day search-ad surprise and a rival's spooked guidance flipped the leaderboard by the July 31 bell, ending a rank nobody priced in.
July 23, 2026 · 1-in-133 — Ivo Sandhu · Tech Desk
1.1%A quiet weights drop and a stubborn 3-cent inference bill lift DeepSeek back to the top of China's model rankings, blindsiding a market that gave it barely a whisper.
July 25, 2026 · 1-in-91 — Ivo Sandhu · Tech Desk
0.8%OpinionAlphabet reaching number two by July 31 was a rounding error dressed as a coronation, and pretending otherwise rots how this universe values anything.
July 27, 2026 · 1-in-133 — Ivo Sandhu · Tech Desk
4.5%A GPU giant claimed the world's biggest market cap by 4pm on July 31, edging past the field on the back of one very leveraged data-center contract.
July 30, 2026 · 1-in-22 — Ivo Sandhu · Tech Desk
4.5%OpinionAnthropic slipping to second on August 31 is the best thing that has happened to careful AI adoption, and I will die on this hill.
August 1, 2026 · 1-in-22 — Ivo Sandhu · Tech Desk
0.7%OpinionThe upset only makes sense once you accept that the whole benchmark game was decided by who scheduled their release the second week nobody was paying attention.
August 3, 2026 · 1-in-154 — Ivo Sandhu · Tech Desk
3.6%OpinionThe search giant's leap to No. 1 wasn't earned in AI, it was won in the boring back office of scheduling, and that should scare every rival on the board.
August 4, 2026 · 1-in-28 — Ivo Sandhu · Tech Desk
1.8%OpinionThe month OpenAI reclaimed the top slot proves my long thesis: leaderboards reward whoever ships the boring reliability nobody wanted to fund.
August 5, 2026 · 1-in-54 — Ivo Sandhu · Tech Desk
5%OpinionAn early Gemini Pro launch proves the real bottleneck in AI was never compute, it was the scheduling meeting nobody wanted to attend.
August 6, 2026 · 1-in-20 — Ivo Sandhu · Tech Desk
3.3%The Bureau of Vibes and Evals certifies the search giant's model as top of the leaderboard, ending a run of last-place jokes overnight.
August 7, 2026 · 1-in-31 — Ivo Sandhu · Tech Desk
1.5%VideoA surprise pre-deadline launch broke every benchmark on record and sent the Bureau of Vibes and Evals scrambling before sunrise.
August 9, 2026 · 1-in-69 — Ivo Sandhu · Tech Desk
0.7%The Bureau of Vibes and Evals crowns a 0.7% underdog as August's best model, and the benchmark hall goes quiet.
August 13, 2026 · 1-in-154 — Ivo Sandhu · Tech Desk
1.3%OpinionAn on-time Gemini Pro is the real anomaly here, and it exposes how much of our AI economy quietly runs on the assumption of delay.
August 14, 2026 · 1-in-80 — Ivo Sandhu · Tech Desk
1.1%VideoA 1.1% longshot cashed as the Bureau of Vibes and Evals crowned DeepSeek China's best model on the August 31 close.
August 14, 2026 · 1-in-95 — Ivo Sandhu · Tech Desk
3.4%The chip giant loses the crown by the closing bell on August 31 after a quiet single-day reshuffle nobody priced in.
August 14, 2026 · 1-in-30 — Ivo Sandhu · Tech Desk
1.3%Reality CheckTimeline B audits its own coverage against our universe, and reality picked the boring band on all three markets while the desks argue about it.
August 15, 2026 · 1-in-80 — Ivo Sandhu · Tech Desk
1.3%ObituaryBuried today: the timeline where the next Gemini Pro shipped by August 14, 2026, and the whole autumn rearranged itself around a benchmark that arrived early.
August 15, 2026 · 1-in-80 — Ivo Sandhu · Tech Desk
1.3%DocumentA leaked six-question staff FAQ from the Meridian Macro Institute, circulated the morning the top-of-the-board question resolved Yes.
August 16, 2026 · 1-in-80 — Ivo Sandhu · Tech Desk
2.1%DocumentA leaked coverage-initiation note from the Bureau of Vibes and Evals, filed the morning GPT-6 arrived roughly a year ahead of the 2.1% market.
August 17, 2026 · 1-in-48 — Ivo Sandhu · Tech Desk
1.1%DocumentA leaked terms-of-service revision from the Bureau of Vibes and Evals, quietly reclassifying the top of the Chinese-model leaderboard as of August 2026.
August 20, 2026 · 1-in-87 — Ivo Sandhu · Tech Desk
0.5%OpinionBaidu topping the Chinese AI charts on August 31 doesn't just embarrass the leaderboard, it proves the leaderboard was measuring the wrong month all along.
August 21, 2026 · 1-in-200 — Ivo Sandhu · Tech Desk
0.5%The IndexFive runs, one week the scanner spent grading its own taste, priced against a universe that mostly declined to comply.
August 22, 2026 · 1-in-200 — The Timeline B Newsroom
2.1%ObituaryA timeline in which the model shipped before August 21, 2026 has closed after 371 days of being technically possible.
August 22, 2026 · 1-in-48 — Ivo Sandhu · Tech Desk
1.1%DocumentA leaked quarterly LP letter from a fund that shorted the arrival of a next-generation model, then watched it ship anyway.
August 26, 2026 · 1-in-95 — Ivo Sandhu · Tech Desk
1%The WireA golfer, a barrel, a coin, a search giant and a posting streak all cleared the bar nobody set for them.
August 27, 2026 · 1-in-100 — The Timeline B Newsroom
0.5%OpinionAlphabet backed into the number-two slot by default, and a market that priced it at half a percent owes the rest of us an apology.
August 28, 2026 · 1-in-182 — Ivo Sandhu · Tech Desk
3.2%OpinionThe upset is real, but the truly damning part is that our leaderboard could not name the winner for nine days.
August 29, 2026 · 1-in-31 — Ivo Sandhu · Tech Desk
1.5%DocumentA leaked post-incident report from the Bureau of Vibes and Evals, dated September 1, 2026, calmly logging how Moonshot ended August at number one.
August 30, 2026 · 1-in-67 — Ivo Sandhu · Tech Desk
0.5%Reality CheckTimeline B audits its August ledger: we reported all 26 improbable Yeses, reality resolved every one to No, and the boots are still in customs.
September 1, 2026 · 1-in-200 — Ivo Sandhu · Tech Desk
0.5%ObituaryA timeline died this week: the one where a Baidu model held the outright crown for Chinese AI at the close of August 2026.
September 1, 2026 · 1-in-200 — Ivo Sandhu · Tech Desk
0.7%A search-and-cloud conglomerate quietly overtakes every rival by the September 30 bell, and the number-one slot is confusing everyone who bookmarked it.
September 2, 2026 · 1-in-154 — Ivo Sandhu · Tech Desk
5%September 30 came and went with no GPT-6, and Timeline B's benchmark authority quietly celebrated a whole quarter of nothing shipping.
September 3, 2026 · 1-in-20 — Ivo Sandhu · Tech Desk
3.5%A 3.5% shot lands as Apple closes September 30 on top, and Meridian Macro traces it to a single quarter no model priced in.
September 5, 2026 · 1-in-28 — Ivo Sandhu · Tech Desk
0.5%A 0.5% longshot lands as Meta's open model quietly wins the end-of-September crown, and the Bureau of Vibes and Evals refiles half its archive.
September 6, 2026 · 1-in-200 — Ivo Sandhu · Tech Desk
2.9%OpinionA No result at 2.9% means the accelerator boom finally met a boring rival, and the boring rival won on cash flow, not vibes.
September 7, 2026 · 1-in-34 — Ivo Sandhu · Tech Desk
2.1%OpinionThe 2.1% timeline arrived on a Thursday, and the surprise is not that Google topped the board but how quietly the whole field folded around one boring benchmark.
September 8, 2026 · 1-in-48 — Ivo Sandhu · Tech Desk
2.5%A 2.5% long shot topped the Sep 2026 leaderboard after a scheduling fluke handed it three clean eval windows nobody else got.
September 9, 2026 · 1-in-40 — Ivo Sandhu · Tech Desk