Happy belated Halloween! I hope you spent the weekend costumed, sugared, and satisfied. My family dressed as the Great Pumpkin from Charlie Brown. I don’t post pictures of my kid online, so you’ll have to kinda picture it with this visual guide.
This week was also a spooky earnings season. All of the big tech companies reported on their progress (stellar), growth (spectacular) and spend on AI datacenters (biblical).
This edition we’ll cover:
OpenAI’s new agreement with Microsoft
My most popular interview yet
How a company in Milan is applying the private equity playbook to Silicon Valley
The new hit R&B artist who just so happens to be AI generated
But first, this edition is brought to you by Composer.
Everyone’s asking ChatGPT to help with investment ideas.
Here’s the problem: it can’t execute. It’s just a research assistant that stops at the keyboard.
Composer is what happens when AI actually crosses the finish line. You describe a strategy. It builds, tests, and trades it at your command. End-to-end. No handoff, no mistakes.
It took the team five years, but they’ve built an AI product that’s executed $20B in trades and has over $200M in managed assets.
The new Trade with AI interface is the unlock. Instead of learning complex tools, you just...talk to it. Share your preferences—tech stocks, ETFs, crypto, whatever. Composer surfaces strategies that match or helps you build custom ones through guided questions.
Every symphony (their term for strategy) includes:
Clear logic explanation
Historical backtest data
Risk statistics
Performance metrics
The AI identifies patterns from thousands of community-built strategies and applies them to your goals. It’s genuinely magical to watch. Five years ago, this tool couldn’t exist. Now it makes executing my trading ideas dead simple.
MY RESEARCH
OpenAI’s new agreement reveals a company under pressure. Microsoft and OpenAI are staying together for the sake of the kids. The companies did a dual press release and podcast tour to assuage any concerns people may have about it. However, most of the press coverage missed the important details, and those details indicate that OpenAI is still operating in early stage startup mode, taking on unfathomable levels of risk for unfathomable levels of reward. You gotta respect it.
Why is the internet bad now? I dig into Cory Doctorow’s theory of “enshittification”—how platforms evolve from vibrant services to bloated, exploitative ecosystems. We walk through the three-act tragedy of digital platforms: lure users with freebies, squeeze them to monetize businesses, then squeeze everyone to satisfy shareholders. I also push back—asking whether convenience truly outweighs privacy and whether every ad network is dead in the water. If you ever wonder why your feed looks worse, your app nags harder, and leaving feels impossible, this essay is for you. You’ll note that there is an interview at the top, but don’t miss the 2000 words of analysis I wrote underneath it! And the interview is great, it is my most viewed one yet.
WHAT MATTERED THIS WEEK?
PUBLICS
Google makes a hundred billion…in a quarter. It is easy to forget just how big and diversified these tech giants are now. Search remains the best business of all time, but the company is getting more and more diversified.
And most of these divisions making billions of dollars are growing double digits.
Google Search & other: Revenue from Google search properties (including traffic from distribution partners) and other Google-owned properties like Gmail, Maps, and Google Play.
YouTube ads: Advertising revenue generated on YouTube properties.
Google Network: Advertising revenue on partner properties that participate in AdMob, AdSense, and Google Ad Manager.
Subscriptions, Platforms & Devices: Revenue from consumer subscriptions (e.g., YouTube TV, YouTube Music & Premium, NFL Sunday Ticket, Google One), platforms (mainly Google Play apps and in-app purchases), devices (primarily Pixel), and other products/services.
Google Cloud: Revenue from enterprise infrastructure and platform services (GCP), Workspace collaboration tools, and other enterprise services.
Other Bets: A bundle of non-Google businesses, with revenue primarily from Waymo and other smaller companies like their work in healthcare.
Doing business this well at this sort of scale is really hard to fathom. Much of previous theories of business strategy that argued against conglomerates and diversification were constructed in the era before networked computers. When you couple Google’s earnings with the success of the rest of big tech, it is worth considering what widespread beliefs of business strategy are now outdated. This shouldn’t be possible, and yet every quarter we witness the biggest tech companies do the impossible and continue to grow. Perhaps this is worth covering in greater depth later this week.
AI LABS
Big tech’s scale justifies OpenAI losing 11.5 billion last quarter. With the new partnership agreement, Microsoft is now giving a greater level of disclosure about OpenAI’s performance. Microsoft had a $3.1 billion hit to its own net income, which given its roughly 27 % equity stake in OpenAI, means that Altman and Co lost nearly $12 billion in a single quarter. That is…quite a lot of money. But as I’ve argued over and over again in this newsletter, OpenAI is going for a ten trillion dollar outcome. Which, like five years ago, would’ve sounded ridiculous, but these big tech earnings make the idea palatable. It is still risky, but the company has to swing this big, otherwise it is likely that the other big tech giants will end up dominating in AI just on the basis of being able to lose more cash for longer.
LAUNCHES
R&B has its first robot hit artist. “How Was I Supposed to Know” by “Xania Monet” debuted at #30 on the Billboard radio airplay charts and has drawn over 5 million streams on Spotify. This success has led to music labels offering multi-million dollar deals to sign the artist. What’s noteworthy is that Xania Monet is the stage name for poet Telisha Jones. She writes the lyrics, plugs those into the AI-music generator Suno, and then her management team works on distributing them.
“How Was I Supposed to Know” originally became a hit on TikTok and social media. The management team behind it pumps out 3-5 Reels a week promoting the music that feature AI generated videos of the “artist” singing. The comments are…a little eerie.
It appears that many of the people don’t even realize this is AI. Beyond the social media strategy, the manager of Monet did an interview on how they focused on radio distribution to increase the artists popularity.
When a product that used to be scarce becomes common and cheap, that merely shifts the competitive vectors elsewhere. In the case of AI, I’ve been arguing for over 3 years that AI will increase distribution costs while reducing production costs. AI is currently great at generating catchy tunes, while it’s terrible at generating lyrics. Therefore, for Xania Monet, the owner being a poet makes a ton of sense. The world already has infinite songs, so how those songs differentiate will fall back to distribution strategy versus the catchiness of the song itself.
How do you feel about this? Is this the equivalent of Bob Dylan using an electric guitar for the first time? Is Monet simply a poet armed with a new type of instrument? Or is it something nefarious? I ask because I still can’t decide what the right answer is myself.
DEAL VIBES
The private equityification of the tech sector continues. Bending Spoons, a Milan-based technology holding company, raised $710 million at an $11 billion pre-money valuation. Simultaneously it added $2.8 billion in debt that it used to purchase AOL (remember that name?) and more deals to come. The company’s operating model is to acquire legacy software brands like Vimeo or Evernote, port their technology to their in-house operations, fire as many people as possible, and then raise prices on existing customers. Not exactly an inspiring mission! Reading the company’s career page is like swimming in platitude soup. Lots of comments about how they promote from within, the great benefits, etc. Nevertheless, this meme is how I characterize this dynamic.
It is important to note that once again, this company is the embodiment of how diversification and scaling is so much easier than it used to be.
$350 million to make the youngest billionaires ever. Mercor is an expert network that large AI labs use to hire people to help train their models. The company is only a few years old and just raised a $350 million round at a $10 billion valuation. To justify the round, the company’s cheerleaders shared this chart a lot on social media.
They are wanting to justify a really big revenue multiple on their growth by comparing themselves to other marketplace businesses. The issue is that these aren’t the same businesses. Airbnb and Uber had to acquire hundreds of thousands of merchants on the supply side, and millions on the demand side of their networks. Mercor needs to source <100,000 people to label data, while only selling to the 15 or so labs that are spending millions on data labeling. Mercor (and all other AI expert networks) suffers from fairly extreme customer concentration risk. I don’t say this as a hater of the firm! These results are incredibly impressive, but in Silicon Valley we too often use false heuristics that rely on the ignorance of the normie. Your business is already great! No need to have false comparisons to make it seem greater.
TASTEMAKER
One quick album recommendation this week! Check out Everybody Scream by Florence and the Machine. You may remember her for the early 2010s hits of Dog Days are Over. This album may be her best work ever. During her last tour, she had a life-threatening ectopic miscarriage onstage, and very nearly died. That scare has tempered this album into something haunting, yet perfect. Her sound has always struggled with being a little bombastic, but here she finally learns to balance the quiet with loud, the screams with the whispers. The high of the album comes in during the synths of track 4 Sympathy Magic. I’ve caught her live twice now, and she’s among the best performers I’ve ever seen. If the tour comes anywhere near you, it’s worth the ticket price.
Have a great week,
Evan
Sponsorships
We are now accepting sponsors for the winter. If you are interested in reaching my audience of 35K+ founders, investors, and senior tech executives, send me an email at team@gettheleverage.com.