Meta Gets The "Accountability" Treatment As AI Chief Blames Zuck For AI Faceplants…

Zuck can’t handle the Wang… 

It appears Alexandr Wang, the 28-year-old Scale AI wunderkind Meta paid $14B for (well, 49% of), is privately annoyed that Mark Zuckerberg won’t get out of his own way. Specifically, Wang thinks Zuck’s micromanaging of Meta’s AI push is… suffocating. Sounds legit. 

(Source: Imgflip) 

In short, back when AI wasn't single-handedly guillotining portfolios left, right, and twice on Sunday… Meta dropped a nice $14B on Scale AI, installed Wang as the face of its AI reboot, and handed him the keys to a secretive internal lab working on a new flagship model codenamed “Avocado.” (No word yet on whether it’s ripe.) Then… Llama 4 rolls out… and spectacularly faceplants. 

In fact, as my parents like to say after I told them I was going to write articles for a living… “Their son Llama was born a disappointment”. Why? Because what it served up was an afterbirth of benchmark lags, and coding performance blue balls. There were even rumors swirling that Meta tried to juice leaderboard results with a custom build. Which definitely isn’t ideal for someone whose already pissed away $70+ billion on the very division Zuck renamed the company for (read: Reality Labs). 

(Source: New York Post) 

Which means inside Meta, fingers started pointing. Some employees blamed bad training data. Others blamed rushed timelines. A few blamed the fact that there are roughly 17 overlapping AI teams all fighting for oxygen like it’s Survivor: Menlo Park. And instead of slowing things down, Zuck did what Zuck does best: He doubled down with a ‘roided up dose of sunk cost fallacy. Hiring blitz’s, compensation packages reportedly brushing up agains $100M. Tens of billions more in capex. Internal deadlines getting shorter. Product launches getting faster. Testing getting… optional. All the works. 

Whereas now, Wang, whose background is more “AI data infrastructure” than “frontier model god,” is now running enormous research orgs inside one of the most politically messy companies in tech. Meanwhile, Zuck is allegedly hovering… deeply involved, deeply opinionated, and deeply allergic to losing to OpenAI. Translation: Meta is having it’s own Mom Talk moment. Because behind the scenes, it’s getting choppy with accountability. For instance, Nat Friedman’s under pressure to ship AI features yesterday. Yann LeCun is reportedly peacing out to do his own thing. Legal chief gone. Revenue chief gone. Hundreds of AI employees laid off… and has somehow been framed as a speed optimization LOL. 

(Source: Imgflip) 

Naturally, investors noticed as Meta shares got smacked in November as fears grew that free cash flow is about to get sacrificed at the altar of “winning AI.” Meta, for its part, says the FT report is overblown… and that everything’s fine. Maybe it is. But when your $14B AI savior is venting to friends about management style, your flagship model disappointed, and your CEO is personally steering the ship from every seat on the bridge… it’s fair to ask: Is Meta building the future of AI… or just stress-testing how much chaos a trillion-dollar company can absorb before something snaps?

Either way, Zuck’s not delegating. And someone is about to catch those hands of his if they aren’t careful. Meaning, keep your eyes on this story and place your bets accordingly, friends. Until next time… 

At the time of publishing, Stocks.News holds positions in Meta as mentioned in the article.