Best of SatPost (First Half of 2026)

PLUS: Meta's AI pivot, Ronaldo, Chloe vs. History.

Thanks for subscribing to SatPost.

Many of you readers are probably starting vacation. This is my last e-mail send before summer break. SatPost will be off for July and I'll be back in your inboxes some time in August (side note: I'm addicted to X, so you can still find me posting on that app).

Today, I'll share my 6 top reads of the first half of 2026 (ICYMI).

And a bunch of links and memes for your summer enjoyment.

6 Top Articles of 2026 (So Far)

I appreciate all you readers.

If I’ve ever made you laugh, say “huh, interesting” or think “wow, Trung has awful takes”…I would love for you to share any of the following articles with friends or family that might like them.

Without further ado, here are the 6 most popular SatPost articles from the first 6 months of 2026 based on views (and general vibes from the e-mail responses I received):

How does Docusign have 7,000 employees?: And other thoughts about B2B SaaS in the age of AI coding agents.
Kirkland Signature and How Jim Sinegal Built Costco's $90B Private-Label Empire: Since 1995, Costco has created over 1,000 Kirkland Signature items. The in-house brand was inspired by a single Forbes article and has turned into a $90B business, 1/3rd of Costco's total sales.
Warren Buffett, Charlie Munger and the Reading Unlock: One Buffett lesson I'll share with my kid: read, read, read.
Tim Cook's Apple: A breakdown of his 15-year CEO run, taking Apple from $350 billion to $4 trillion: The Good, The Bad and the Apple Intelligence.
The Sphere's $4B Business, explained: The eye-catching Las Vegas attraction cost James Dolan $2.3B to build. After an uncertain start, the recent success of "The Wizard of Oz" has created a clear playbook for global domination.
SpaceX's AI IPO: Notes on SpaceX's S-1 document (and Elon's bid for a $1.75 trillion valuation)

This issue of SatPost is brought to you by Bearly AI

Is your team looking for a platform to manage multiple AI models?

One that is built for your company's specific data, UX and business needs?

Try Bearly AI, which has been working with enterprise clients to connect their organizations with leading LLMs (Claude, ChatGPT, Grok, Gemini, Kimi, DeepSeek), including customizable UXs, API management, budgeting dashboards and complete data privacy.

Let's build something for your team.

Book A Demo

Links and Memes

Meta's AI pivot: Bloomberg reported that Zuck Daddy Flex is planning to turn Meta into a provider of AI cloud compute. Meta pumped +10% on the news but the rest of the AI supply chain sold off as many investors felt this was a sign that Zuck had overspent on capex.

SemiAnalysis believes this is a misunderstanding of Meta's strategic position and lays out how Zuck has a lot of optionality with his capex and expects Meta to keep spending:

Of course, this naturally raises the questions of what Meta will do with this compute, and whether they’re going to flood the market with all of this supply if they turn into a Neocloud. Broadly speaking we see four major high-value use-cases, which are all differentied and very different relative to what traditional Neoclouds do:

1. Frontier AI Models: Meta has NOT given up on training frontier models. The bulk of incremental capacity still goes to Meta Superintelligence Labs, and we think the team is currently excited about their progress. A follow-up report will dive into MSL, their unique data strategy, and discuss their chances of catching up with Anthropic and OpenAI. Of course, our Tokenomics Model subscribers already know our takes and have access to all of this real-time.

2. RecSys: We believe Meta thinks they can scale up Ads recommendation systems by >10x in complexity to accelerate revenue growth. That requires both inference & training compute for their RecSys models. They can also do more generative targeted ads.

3. (SemiAnalysis Exclusive) We believe that Meta is in final talks with Anthropic to get access to private instances of Claude. This would be akin to Bedrock, Foundry, Vertex from other hyperscalers (read our deep dive here). There are multiple use-cases for Meta, ranging from internal usage, to building the premier Sales & Marketing SaaS powered by Frontier AI Agents. We expect Meta to launch a token-as-a-service endpoint and increasingly move up the stack, leveraging its network and distribution. Initially it will be their own models externally and Anthropic internaly, but over time we believe they will serve Anthropic and OpenAI models externally

4. We expect Meta to strike a few “SpaceX-type” deals. Elon is a sales genius, and he created a brand new market segment: large-scale on-demand compute at a huge pricing premium. We think Meta wants in, but selectively. After all, just a couple hundred MWs can already drive >$10B of yearly revenue! We expect a ten billion dollar Anthropic deal to kick off the flywheel.

This high optionality, with four high-value-add options, makes it easy for Meta to keep aggressively contracting compute. Meta Superintelligence remains the core engine, but if it doesn’t work out, there are many high-margin alternatives to sell compute. It is essentially a CFO’s dream and makes it very easy to go all-in on compute – we bet Susan did a 180° flip when she saw the pricing of SpaceX compute deals! Meta won’t be a normal bare-metal IaaS vendor with ~30% gross margins – all its options are high value, and enable to easily afford paying a margin to other Neoclouds in order to accelerate their fleet buildout - even if MSL doesn’t work out.

Some other links for your weekend consumption:

The American Revolution: Happy July 4th to all my readers who celebrate. Ken Burn's The American Revolution was the last full TV series that I watched and here are my notes on the 6-part series released in the lead-up to America's 250th birthday.
The New Yorker's profile of Erling Haaland...has an incredible opening few sentences: He sometimes exercises by chopping wood in the forest. He consumes 6,000 calories a day. After training sessions, he drinks raw milk. He owns a tax-sheltered investment company in Luxembourg named Pillage. He bought an edition of the “Heimskringla,” a 13th-century Old Norse saga, for $130,000—then donated it to his local library because, he explained, “I’ve never been much of a reader.”
The Rest Is History has a solid series on national anthems...including Brazil, Germany, USA, Netherland and UK. The back story on South Africa's anthem was fascinating.
Ronaldo tops Instagram with 670 million followers...and his latest post — with his team on the balcony of a Toronto hotel waving to Portugal fans after they beat Croatia — is a reminder of how famous this dude is. The last time he was able to walk the streets like a normal person was when he put on a padded fatsuit, wig and mustache to clown people with a soccer ball in central Madrid.

Chloe vs. History is the best use of AI-generated video. A 20-something "girl" travels to different historical periods in a Vlogger style with 2026 internet vernacular. I put "girl" in air quotes because its all AI-generated and legit fooled me. She blew up on Instagram but now has >10 minute episodes in during the Ice Age, Tudor England, Ancient Egypt and more. Embarrassingly, I thought it was a creator who just superimposed herself on AI background...not realizing she herself was AI. Legit googled "what school did Chloe go to?" after watching two episodes. I'm an idiot. Anyways, Chloe is the creation of a 32-year old British filmmaker named Jonathan Laramy...and here is a solid podcast with him explaining his process. You got me Jonathan!

…and them wild posts (including a great running joke on the World Cup's VAR offside technology):