Bullpen.News From Edwin Dorsey
Get Daily Recap Feedback

Friday, October 30, 2026

Back to today

From X

@convequity

Convequity on X

The Artificial Analysis (AA) Intelligence Index update is interesting, and the right reading is more nuanced than its one headline number. AA is an amateur score, aimed mostly at non-specialists. It is not the most authoritative benchmark, the most professional one, or the most accurate one, partly because of how AA sets up its own tests. On FrontierCode-AA, GPT-6 Luna (medium) is the language model that grades the answers of frontier models including GPT-6 Astra, Claude Fable and Claude Opus. AA uses that judge on FrontierCode 1.1 and 2.0. On FrontierCode 2.0, Luna scores about 25, while Claude Opus 5.5 scores about 46, GPT-6 Astra about 42 and Claude Fable 5.1 about 36. So the judge scores below each of the frontier models it grades. People still use AA because it turns "which model is better" into one number. That number is easy to quote, so labs put it in launch announcements and buyers use it to compare models quickly. Many labs intentionally overfit: they train a model on the answers and the tests from the benchmarks that made up AA's score, so it scores well on those tests without getting better at the work. About 60% or more of the benchmarks that constituted that score were public. On a public test, the questions and answers are published, so a lab can put them into training data and the model can memorize them. The published score then rises because the model has already seen that test, and that score is the number the market will quote. Labs under pressure to show a strong result overfit those public tests hard. The result is high scores on the tests and low scores in actual use. Real work brings new code, documents and problems the model has never seen, so memorized test items do not carry over. So far, $META's Muse, $GOOGL's Gemini and Moonshot's Kimi seem to be the most aggressive. Meta and Google face the most pressure to show they are back on top. They gamed the tests heavily, and their actual performance remains weak. A performance claim is now easy to check in actual use. Right after Opus 5.5 shipped, Anthropic's token use jumped and many people moved back to Claude Code, while Muse Code and Google Antigravity remain weak. Token spend and the coding product people actually open show what customers do with their own work, and a model trained on a public test cannot produce that shift. If Muse and Gemini had the lead their test scores implied, use would have moved toward Muse Code and Antigravity, and it did not. Benchmarking remains one of everyone's biggest problems. Weak benchmarks saturate. A weak test has a ceiling, and once the top models all reach the maximum score, the ranking stops separating them, even if one model is still better at the work. Strong benchmarks with public questions get gamed, because labs can train on the published questions. Closed benchmarks keep the questions hidden, which makes them harder to train on. The cost is weak transparency and reproducibility, and they are harder to treat as independent. A buyer cannot see the items, rerun the test from those questions, or check that the lab and the scorer stayed apart, so the result is less trustworthy. AA's new v5 index is 60% private or held-out, meaning the questions in that 60% are kept out of public view, and this is the first time a majority of the index is private. That private share is a different figure from the 60% or more of the earlier tests that were public. Even with that cost, I think AA made two right moves. It now uses more closed benchmarks, and it dropped its own poorly designed special benchmarks: GDPval-AA v2.1, SciCode and AA-LCR v1.1. Apart from the single headline number, the leaders on each test are still informative: Real code-repository development (FrontierCode): Claude Opus 5.5, 46% on FrontierCode 2.0 at max effort, the same result as above Terminal engineering tasks (Terminal-Bench): Kimi K3.1 Pro, 67% on Terminal-Bench 4.0 Terminal scientific research (Terminal-Bench-Science): GPT-6 Astra, 63% at max effort Complex physics research questions (CritPt): GPT-6 Astra, 72% on the private hard subset at high effort Enterprise knowledge work (AA-Briefcase): Claude Sonnet 5.5, 1,812 Elo Enterprise SaaS automation (AutomationBench): Gemini 4 Argon, 78% Complex PDF understanding (GDP.pdf): Muse Wonder Full, 33% all-pass HLE-Diamond, the refined hard-exam set: Claude Opus 5.5, 68% at max effort Knowledge reliability (AA-Omniscience): Claude Opus 5.5, 48 on AA-Omniscience v1.1
@BlueDuckCap

BDC on X

Factor and sector rotations off of day to day and often conflicting data points have been a hallmark of the 2026 market. Not sure if its algos or just heavy retail participation or both. But I would say this: stop reacting to the daily data points - especially during earnings. It's stupid and a recipe for getting chopped into oblivion. While earnings reports are useful for directional trends on sectors, every company is different. Obviously. And different things impact quarters. To wit - its Thursday. Sunday night: AI trade over. Monday after Anthropic S1 out: Oh there's a healthcare TAM here and inherent operating leverage on compute, buy AI! Tuesday: AI stable, sell $SPOT! Weds: Send AI Weds PM: $GOOGL sucks, puke it, its over. Buy $META (why? idk). $NOW sucks puke it. Capex peak in sight (in line with consensus) puke AI. $MSFT good, buy it. Thurs: AI over, software over buy $SPOT. $MSFT good but its AI and software, sell it. Try and have an investment view anchored in what you believe to be confirming data points you expect to see in coming weeks, months, years. Hedge tactically as the data points inevitably aren't linear. Barbell core investments with tactical trades to hedge and take advantage of the rotational chop. Such trades are best in highly liquid securities. But if you trade like a bipolar lunatic - much like the market has traded - you will not do well.

Investor Letters

All letters

SNY Earnings Review

Sanofi raised FY26 guidance after Q3 sales grew 14.3% at constant currency, led by 34.8% Dupixent growth; vaccines declined 6.6%.

SNY $108B
Cestrian Capital Research 1 click

TSOH Weekly Roundup (10/30/2026)

Markel spent $315 million repurchasing shares in Q3 as Thrive agreed to buy ADT for $7.6 billion and Hasbro cited rising adult toy demand.

MKL $23B
TSOH Investment Research 1 click

AMZN Q3 FY12/26 Earnings Review

Amazon reported $201.9 billion Q3 revenue and 39% AWS growth, while $220 billion in 2026 capex drove trailing free cash flow to a $42.4…

AMZN $2.6T
Cestrian Capital Research 3 clicks
SAP · long

Sap Se

SAP shares are down 45% since February 2025 as AI disruption fears create a long thesis on its risk-reward.

SAP $272B
Value Investors Club 2 clicks

Another Haircut with Avy the Barber

SoFi’s stock case hinges on interest rates, tangible book value and private-credit loan sales after its $1.5 billion equity offering.

SOFI $22B
WHAT ARE WE DOING? “Contrarians At The Gate” 3 clicks

Intel Q3 2026: Dead Inside

Intel’s Q3 2026 datacenter CPU shortage reflects wafer-start planning failures, not substrate constraints, after demand shifted from client…

INTC $348B
Irrational Analysis 2 clicks

Twilio CEO: ‘Buy the dip!’

Twilio CEO Khozema Shipchandler urged investors to “buy the dip” after earnings sent the software maker’s shares down 17% after hours.

TWLO $39B
The Technology Letter 2 clicks

Publications we follow

Every publication on this site, hand-picked.

129 publications

Newsletters 109

YouTube 4

Blogs 15

Seeking Alpha 67 of 268

Nominate a publication

X accounts we follow

Every account on this site, hand-picked.

91 accounts
Nominate an account

Investor letters we follow

190 firms

Sectors

Ordered by how often our editor reaches for them.