Happy Wednesday. Mark Zuckerberg might just write one of the biggest checks of his career— a $14.8 billion bid for 49 % of Scale AI. If the deal closes, 28-year-old Scale AI CEO Alexandr Wang will moon-walk from his San Francisco HQ to Menlo Park with both a Meta badge and a DoD pass in his pocket—hot on the heels of Scale’s new “Thunderforge” contract to turbo-charge military planning with LLMs.

 Deals

L’Oreal to Buy Majority Stake in British Skincare Brand Medik8

Kering Buys Italian Manufacturer Lenti in Eyewear Push

Qualcomm to Buy Alphawave IP for $2.4 Billion

Industry News

Yale Is Rushing to Sell Billions in Private Equity Investments

Blackstone plots $500bn expansion in Europe

World

US and China agree framework to restore trade truce

American tariffs are starting to hammer Chinese exporters

Zuckerberg is diving headfirst—once again—into the AI pool, scooping up nearly half of Scale AI in a deal that places the valuation at an eye-watering $28 billion.

Though deals like this often sputter right on the verge of completion, people close to it indicate an official announcement could come as soon as Wednesday.

Now, absorbing a commanding 49 percent isn't just idle investment; it's part of Meta's grand strategy to assemble an enclosed "superintelligence" lab.

The goal here? Eclipse rivals like OpenAI, Anthropic, and Google—companies racing alongside Meta in what's rapidly shaping up to be tech's ultimate arms race: building models smarter than the human mind itself. Ambition seems to fall short as an accurate descriptor here—audacious might come closer.

But here's where things twist dramatically...

Llama 4, Meta’s highly anticipated language model, turned out to be a significant letdown in crucial independent reasoning and coding tests.

I can't help but sense frustration under the surface. Critics didn't exactly spare Mark's team, contrasting Meta’s bruising misstep against Google, OpenAI, and Anthropic’s newer generations of reasoning-focused AI that systematically untangle problems rather elegantly.

Digesting such setbacks can't be pleasant—especially now, as poised Chinese open-source competitors like DeepSeek churn out similarly powerful models on shoestring budgets, evidently determined to rattle Meta's cage.

For context, Meta’s AI investments don't exactly lack heft. With the organization floating toward a $2 trillion market cap, resources aren't the issue—direction may be.

Time and again, executives—including Joelle Pineau, who quietly exited back in April—shuffled or departed, leaving Meta's AI initiatives stuck on repeat in a perpetual pivot. You can almost sense the anxiety rippling through their ranks: Is Zuckerberg's obsession with generative AI about to pay off, or merely compound recent missteps?

And now Alexandr Wang, the ambitious 28-year-old "paper billionaire" behind Scale AI, steps into this simmering cauldron, reportedly joining Meta’s new superintelligence endeavor.

Known for sharp networking chops rather than managerial finesse or deep research contributions—The whispers are that Wang excels more at selling vision than executing it—he’s certainly a striking wildcard.

Whether this billion-dollar gamble wins big or fizzles spectacularly, Wang's jump to Meta guarantees fresh ripples across Silicon Valley.

People familiar with Scale AI mention growing uncertainty...

What does the deal mean for Scale's staff, projects, and the original IPO dreams Wang appeared keen to realize?

Jason Droege, freshly plucked from Uber Eats a year ago, was reportedly being groomed for chief executive. But now that role—and Scale’s future direction—stand shrouded in ambiguity.

Wang’s firm had also lately shifted gears, expanding revenue streams beyond its core manual data-labeling operations, trying its hand at tailored enterprise apps and lucrative government contracts. With regulators sniffing closely around tech heavyweights—Google and Microsoft ran into their fair share of antitrust headaches after similar deals—it’s impossible to ignore the increases in creative financial maneuvering now standard practice in Silicon Valley.

In a tech landscape where Microsoft threw $650 million at Mustafa Suleyman's Inflection team and Google shelled out potent billions for Character AI partnerships, Meta's latest venture embodies Silicon Valley’s frenzied pursuit of the next great leap, regulated or not.

I can't shake the feeling we’re poised on yet another rollercoaster moment—waiting to see whether Zuckerberg’s latest billion-dollar bet on superintelligence conquers new horizons or lands as an expensive, humbling stumble.

If I were a betting man (I am), I'd say buckle up—things are about to get very, very interesting.

See you next week,

— Matt

P.S. If you enjoy my emails, move them to your primary inbox and let me know by sending a reply or clicking the poll below. I read every response.