Bullpen.News From Edwin Dorsey
Get Daily Recap Feedback

NASDAQ

Meta Platforms, Inc.

META

218 stories

$1.6T Market cap · 2026-10-28

@convequity

Convequity on X

Thoughts on Elon's recent neocloud comments "Neocloud" has always been a weak label. These businesses aren't really clouds, and they aren't particularly new. The more accurate term is GPU VPS. Most of them rent bare-metal GPUs. They generally don't offer redundancy across three availability zones, and they don't sell hundreds of platform services on top of the hardware the way a true cloud does. That puts them closer to VPS providers than to cloud platforms. CPU history shows the difference. CPU cloud concentrated among three to five players because platform services lock customers in and drive a demand flywheel. Customers use more of the stack over time, rather than only renting the hardware. That is how AWS-style clouds kept high margins and strong moats. CPU VPS went the other way: many providers selling essentially the same product, in a brutally competitive, commoditized market. If GPU supply stops being constrained by chips or energy, GPU VPS starts to look like CPU VPS, with gross margins of something like 50%. Better marketing and operations can lift margins and grow revenue, but they won't push gross margins to 75-80%. That is the long-term risk Musk pointed to when he predicted most neoclouds would go out of business. Most of these names are VPS businesses earning cloud-like profits and valuations from a temporary imbalance between supply and demand. $CRWV is part of that group. CoreWeave and $NBIS are also the two furthest from a pure GPU VPS, because both already sell platform services on top of the GPUs. CoreWeave is still the most exposed on financing. It pushed hard to maximize margins on each contract and took on interest-rate risk against fixed contracted revenue. Its counterparties, OpenAI, $MSFT and $META, may have played those deals more intelligently. So CRWV is the most vulnerable name on its balance sheet, yet one of the least commodity-like on product. That gap is where the return potential sits for investors willing to stick with CRWV.
@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.

Meta: A Clearer Roadmap

Meta's non-cancelable commitments reached $455 billion as FY26 capex approaches $140 billion, driving expense growth and 700-basis-point…

META $1.6T
TSOH Investment Research 3 clicks

Meta Earnings, Muse and the Frontier, Musebook

Meta’s earnings hinge on whether investors back Zuckerberg’s Muse vision and the compute spending required to build it.

META $1.6T
Stratechery by Ben Thompson – On the business, strategy, and impact of technology. 1 click
@ParadisLabs 1 click

Paradis on X

A lot of green in AI land today from $NBIS +11.2% to $CRDO +10.8%. Over the past few weeks, financing has been the main pressure point but it seems like we've made some solid progress this week. - $META raised $97.5B, with $60B of that coming through an ATM starting in January. - $SPCX signed a $40B deal for GPU equipment. - $AVGO set up a $63B SPV to fund custom chips for Anthropic, who also eased some fears with their S-1. Feels like the rapid de-grossing over the last few weeks has mostly run its course, however according to JPM, volumes still remain light. Interestingly, GS say that there's evidence today of "increased offense being played across the AI complex". While this is still minimal, it's a notable uptick vs. the past few weeks where risk was near zero. I wonder whether the market decides to pay attention to the upcoming hurdles, or whether it keeps riding the wall of worry higher. I personally still land on some short-term choppiness when you consider how bruised investors have gotten since July + hurdles in the next few days (hyperscaler earnings, FOMC, midterms).
@aleabitoreddit 4 clicks

Serenity on X

I've always defended X as the only platform to be on for financial discourse. But it's getting harder to when the owner constantly amplifies narratives of collapse… of companies whose interests conflict with his own at $SPCX or $TSLA. Would people be comfortable if $NVDA Jensen Huang owned $META? And then the timelines were dominated by his amplified narratives of $AVGO failing? I can't see how that would be healthy for grassroots discussion.
@aleabitoreddit 11 clicks

Serenity on X

So back on your Neoclouds like $NBIS + memory read through... From $AMZN CEO + A16z interview. Q: On memory shortages (eg. $MU / SK Hynix / Samsung): A: "There's only so much memory, and I don't think that that shortage ends in 2027." Q: On AWS capacity: A: "A good chunk of 28 is sold out", "largely sold out through 2027." Q: On capex ROI: A: Amazon had ​said it would spend about $220 billion in capex for 2026. But Jassy said the company "could invest a lot more ​and it wouldn't be a bad use of ​capital" This is probably validating economics of Neoclouds model since this comes from a pretty stingy hyperscaler like Amazon... that only serves free bananas to their employees. Unlike trusting $META / Zuck where you the guy throws all you can eat cruise ship type buffets at every employee. - AWS claims approximately 30-40% better price-performance than comparable AI accelerators. Then went on to say Annapurna was "one of the very best acquisitions we've done in Amazon and the history of the company, if not the best." "We spun out a separate team in Annapurna to build an AI accelerator" Just to give a bump to your AI custom silicon type private companies (since others like $AAPL were likely going on shopping sprees). TLDR: - Not really too much groundbreaking, just incremental validation of what was said previous earnings. - Broader shortages (like memory) won't end in 2027 - Always positive to hear from $AMZN on insatiable demand + ROI given atrocious Neocloud sentiment right now
@convequity 2 clicks

Convequity on X

$META's Muse progress confirms our view that Meta owns the essential assets for frontier AI, and it is progressing better than most people believe. Zuckerberg called Muse Wonder Meta's first frontier model on the 21 October Q3 2026 call. More than 10 million people set up Muse in about six weeks, and shopping is the largest use case. We expected Meta to come back, and we are delighted to see it back among the frontier labs. Two things concern us. 1. Meta appears to have recognized it will not catch up on frontier enterprise use cases, coding above all. So it is differentiating on consumer. In our view that is a double-edged sword. Consumer is where Meta's assets are strongest. But we think the next big shift is silicon-based consumption squeezing out carbon-based consumption: spending by AI agents replacing spending by humans. Humans have finite attention and a limited addressable market, so human demand grows roughly linearly. AI agents have a near-unlimited spending demand, and we think it is about to go exponential. A consumer-first AI is built for the side that grows linearly. That is a longer-term concern for Meta. 2. We are disappointed that Muse is only available in the US, not in the rest of the world. That is cautious big-company behavior, closer to Google than to SpaceX. Our guess is that the legal warnings and threats Meta keeps receiving have made it extra careful on the legal side. That caution has a cost: it handicaps Muse against the other leading frontier apps.
@convequity 1 click

Convequity on X

Banks are right, for now, to block third-party AI agents from taking action inside consumer bank accounts. An agent paying with a capped card, as $META's Muse does, is a different and narrower permission. Axios reported this week that major U.S. banks are finding unapproved third-party agents inside customer accounts, most let in with the customer's own login. Some tried actions that looked independent of what the customer asked. $MS is blocking such agents, per Axios sources. $JPM, $BAC and $C are restricting them. Banks are slow to change how a deposit account works for a reason: the duty runs to every customer and to the bank's own systems, not only to the early adopter who opted in. An agent logged in with real customer credentials puts fraud, liability and security risk on the bank, and that risk reaches customers who never opted in. The bank is right to keep that action out of the account for now, even if some customers want it. An agent holding the login can see the whole account and send money out. An accepted wire is generally final and sits outside Regulation E, the consumer rule for electronic transfers. An ACH transfer the account sends has no chargeback. Getting it back depends on the receiving bank agreeing. Tell an agent to pay, and an ACH payment inside that instruction can count as authorized under that rule. Money lost inside that grant is the customer's. Jamie Dimon called this a liability transfer: "If people are using them, they should understand the risk." That settles the customer's loss, not the bank's exposure. Muse pays through Stripe Link or a single-use virtual card capped at the total the buyer approves. The customer can dispute a wrong charge. The issuer can cancel that card so the next attempt fails. The agent never touches the deposit account. Agentic commerce has to prove itself on the capped card before that access reaches the deposit account.

Investor Letters

Letters that discuss META.

Every letter that discusses META

Tickers on this page

Companies the stories above are also about, besides META.