GM. This is Milk Road AI, covering the AI arms race, the capital fueling it, and the bets that could reshape the global economy.
Here’s what we’ve got for you today:
- ✍️ Meta’s all-in push to build an AI empire.
- 🎙️ The Milk Road AI Show: The AI Investing Playbook Nobody Talks About w/ Ronnie.
- 🍪 OpenAI bets on audio as the next AI interface.
Launching The Energy Network on Solana, Fuse Energy has just secured $70M in Series B funding. Discover the future of energy now.
Prices as of 10:20 AM ET.
INSIDE ZUCKERBERG’S PLAN TO TURN META INTO AN EMPIRE
There is a famous scene in Breaking Bad where Walter White is asked a simple question:
"Are you in the meth business or the money business?"
He stares back with chilling clarity and says:
"I'm in the empire business."
For nearly 20 years, we thought Mark Zuckerberg was in the social media business.
We thought his goal was just to connect you with your high school friends, help you stalk your exes (guilty! Sorry, Maria), and sell you ads for orthopedic sneakers.
But entering 2026, he is playing a completely different ball game.
Zuckerberg is trying to build the operating system for the global economy.
And he is spending a mind-melting $100+ billion in 2026 to do it.
Talk about a mid-life crisis.
To be fair, Zuck isn't the only one lighting money on fire.
The entire neighborhood is being renovated.
Just look at the projections for 2026 from Morgan Stanley:
- Microsoft: ~$147 billion.
- Amazon: ~$140 billion.
- Google: ~$135 billion.
- Meta: ~$115 billion.
Combined, these tech giants are dropping nearly $600 billion in 2026 and are on track to hit $700 billion in 2027 on infrastructure.
But Meta is different.
Amazon, Microsoft, and Google are "Landlords".
They build data centers to rent to other companies (AWS, Azure, Google Cloud).
They have to spend this money to keep the lights on for their tenants.
Meta is the only one essentially building a fortress just for itself.
They don't have a massive Cloud business to pay the bills.
Every single one of those $115 billion dollars is being bet on their own products. Their own AI, their own ads, their own reality.
It is the biggest "all-in" bet in business history. And frankly, it looks terrifying on paper.
But Zuck has found a glitch in the matrix.
(And no, it’s not just the fact that he still doesn't seem to blink.)
Here is the master plan, explained simply.
The cash cow (using AI to print money)
While the other hyperscalers are "landlords" waiting for rent checks, Meta found a way to monetize their data centers immediately.
They rebuilt their entire ad system using new AI models like "Lattice" and "Andromeda".
They sound like names General Aladeen from Dictator panic-invented while staring at a fire extinguisher.
But here is what they are actually doing:
- Lattice predicts exactly which ad you will click.
- Andromeda boosted Facebook ad quality 14% in a quarter.
Their automated ad tools (Advantage+) are now generating $60 billion a year on autopilot.
This creates a terrifying loop for their competitors:
Better AI = Smarter ads = More profit = More GPUs to build even better AI.
The brains (owning the AI supply chain)
But money isn't enough. You need brains.
And in early 2025, Meta realized its brain was officially in trouble.
Their flagship model, Llama 4 Behemoth, was a flop.
It was big and expensive, but it had the reasoning skills of a goldfish.
It had hit the "Data Wall" (the digital equivalent of finishing the entire internet and still being hungry for more).
Zuck realized he couldn't just keep scraping Reddit threads and call it "learning".
He needed the good stuff, the "organic, grass-fed" data.
So, in true Empire fashion, he bought the supply chain.
Meta dropped $15 billion to snag a 49% stake in Scale AI and kidnapped (I mean hired) their CEO, Alexandr Wang, to run a new "Superintelligence" lab.
Why the massive price tag? Because Scale AI is essentially a high-speed refinery for "dirty" data.
Most AI companies have mountains of raw data, but it’s basically digital sludge, filled with typos and Reddit arguments.
Scale employs a literal army of PhDs and mathematicians to act as human tutors.
They grade the AI's homework until it actually learns how to think rather than just mimic.
By bringing Wang in-house, Zuck got the ultimate "God View" of the industry.
You see, because almost every major AI lab (OpenAI, Google, even the Pentagon) used Scale AI to refine their data, Wang had been reading his competitors' diaries for years.
He knew exactly what data OpenAI was buying, what secret features Google was cooking up, and most importantly, exactly where they were all failing.
Meta essentially stopped guessing where the finish line was and hired the guy who was building the track.
So now Meta has the money and the brains.
But a genius brain without limbs is just a very expensive calculator.
And to win the war, you need the hands (guess what they just bought).
FUSE ENERGY RAISES $70M AT A $5B VALUATION
Is this the most legit energy company to ever enter crypto?
Fuse Energy is a $400M ARR utility powering 200,000+ homes, today announcing a $70M Series B at a blockbuster $5B valuation.
This comes after the recent beta launch of The Energy Network, a new digital layer engineered to scale our grids and save billions in costs.
And now, it’s just building its momentum:
- Today raised $70M in Series B led by Lowercarbon and Balderton.
- Now valued at $5B.
- Launched beta on Solana.
- Received landmark no-action letter from the SEC last month.
- Planning listings for early 2026.
A new foundation for the grid is coming.
Check out their announcement here and follow Fuse on X for updates.
INSIDE ZUCKERBERG’S PLAN TO TURN META INTO AN EMPIRE (P2)
Last week, Meta dropped roughly $2 billion to acquire Manus, a startup whose name literally means "Hand" in Latin.
While OpenAI and Google are frantically trying to build agents, Manus is arguably the leader.
According to the latest Remote Labor Index (a benchmark that tracks how many real-world human tasks an AI can fully automate), Manus holds a 2.5% automation rate, beating out Grok 4, Sonnet 4.5, and even GPT-5.
I know what you are thinking: "Manus is at 2.5% and Grok is at 2.1%... that’s not exactly a blowout".
It took the other guys billions in R&D to reach that level.
Meta realized they could never just buy OpenAI or Google (neither are for sale), so they did the next best thing:
They went shopping for the best independent tech on the market and bought the whole team.
This is where all the pieces start to fit together.
Imagine you run an Instagram shop selling vintage sneakers.
Instead of hiring an assistant to handle orders, you activate your Meta Business Agent (powered by Manus):
"Manage all my shipping labels and inventory for the week."
It logs into your Shopify backend, checks the new orders, navigates to the FedEx portal, buys the shipping labels, and updates the tracking numbers automatically.
It literally does the job while you sleep.
Or imagine the consumer endgame with the Meta Glasses:
You are walking down the street wearing your Ray-Ban Metas. You see a poster for a concert you want to see.
You don't pull out your phone. You just look at it and say:
"Hey Meta, get me two tickets for that."
Because Manus has "Hands", it sees the poster, scans the QR code, navigates the Ticketmaster queue (fighting the CAPTCHAs for you), selects two seats, and buys them before you’ve even walked past the venue.
So Zuck is basically building the WeChat of the West, a Super App that handles your entire life.
The moat (the attention monopoly)
But frankly, even if the agents are mediocre, Meta still wins.
Why? Because of the moat (competitive edge).
There are 3.54 billion users using Meta apps every single day.
That is nearly half the planet.
This creates a distribution moat that is basically impossible to cross:
- OpenAI has the smartest chatbot, but they have to beg you to download their app.
- Meta owns the screen you watch 50 times a day.
In an AI-driven advertising ecosystem, feedback loops are everything.
Every time you like a Reel, reply to a WhatsApp message, or click an ad, you are training Meta's AI.
Meta doesn't just have the best data. They have the most data.
And in the AI wars, he who holds the data and distribution holds the crown.
And unlike OpenAI, this moat prints cash.
Speaking of which, let’s take a look at Meta’s income:
Whoa, why did profits fall off a cliff in Q3? That is what we in the biz call a "nothingburger".
That drop is entirely due to the "One Big Beautiful Bill Act", which slapped Meta with a nearly $16 billion one-time tax bill.
It looks ugly on a chart, but look under the hood: Without that one-time charge, Meta's profit would have been a staggering $18.64 billion.
The underlying business is so strong that it can eat a $16 billion government fine for breakfast and still be profitable.
The verdict: Empire or bust
We obviously want this "Empire" plan to work.
We want the agents, the glasses, and the superintelligence or whatever else Zuck is secretly planning.
But what if it fails? What if Zuck burns $115 billion and the AI agents are dumb?
The infrastructure is still valuable.
Zuck even said it during the Q3 call himself: "The very worst case would be that we effectively have just pre-built [infrastructure] for a couple of years... we'd grow into that and use it over time".
Translation: If the robots fail, he will just point them back at making Instagram ads even scarier and Reels even more addictive.
It’s the ultimate hedge.
And that’s why Meta is a good stock to watch in 2026.
They are making solid, coherent moves in AI.
But I am still waiting for confirmation to see how all this crystallizes.
Why? Acquiring talent is easy but making it work inside a massive organization is the hard part.
Look, love him or hate him, you have to respect the guts.
Most CEOs are terrified of missing a quarterly earnings target by a penny.
Zuck is out here burning billions because he’s looking at the scoreboard in 2035, not next Tuesday.
He’s playing a completely different game than anyone.
He’s not in the meth business. He’s in the empire business.
Alright, that’s it for this edition of Milk Road AI. We want to hear from you.
Reply to this email with your vote:
Is Zuck crazy or a genius?
- Genius: He’s playing 4D chess while everyone else plays checkers.
- Crazy: $115 billion is going to burn the company down.
- I just want my news feed back.
THE AI INVESTING PLAYBOOK 🔍
This past Friday, we sat down with Ronnie, from The Ronnie V Show, to talk about how he’s rotating capital into misunderstood AI and infrastructure opportunities most investors are missing.
Here’s what you’ll hear:
- Why Ronnie sold Google after its big run and reallocated into Adobe instead.
- How tokenized AI usage can turn software like Adobe and Zeta into recurring revenue machines.
- The under-the-radar AI infrastructure plays in energy, robotics, and healthcare.
- Ronnie’s highest-conviction multi-year pick heading into 2026.
Don’t sleep on this one 👇
YouTube | Spotify | Apple Podcasts
Bridge helps businesses send, store, accept, and launch stablecoins instantly. Learn how Bridge is powering stablecoin-backed money movement.
Chainlink is the industry-standard oracle platform bringing the capital markets onchain, and the market leader powering the majority of decentralized finance. Learn more about Chainlink.
Arc is a new Layer-1 blockchain designed to support real-world economic activity and it’s now live in testnet. Dive into Arc and start experimenting.
BITE-SIZED COOKIES FOR THE ROAD 🍪
OpenAI is betting big on audio-first AI as it prepares an audio-focused personal device. The push reflects a broader tech shift toward voice replacing screens as the primary interface.
Nvidia is in talks with TSMC to ramp up H200 chip production as China demand surges. Orders reportedly exceed Nvidia’s current inventory by a wide margin.
OpenAI is paying employees about $1.5 million each in stock-based compensation. The figure is the highest ever for a major tech startup.
MILKY MEMES 🤣
ROADIE REVIEW OF THE DAY 🥛
