According to Gizmodo, Meta’s AI ambitions appear to be in serious trouble despite massive spending. The company committed to pouring billions into AI infrastructure in front of the President and spent $14 billion to acquire Scale AI in June, bringing CEO Alexandr Wang under Meta’s umbrella. They followed this with multi-million dollar contracts for top AI researchers, including some $100 million per year paydays that could make recipients quarter-billionaires. Since launching its Superintelligence Lab in June, Meta broke it into four teams in August, then cut around 600 AI team members in October while delaying the underwhelming Llama 4 Behemoth model. Meanwhile, Meta’s AI app reached just 2.7 million daily active users after introducing Vibes, barely cracking the top 50 of most used web apps according to Andreesen Horowitz’s ranking.
Money Pit Problems
Here’s the thing about throwing money at problems – it only works if you actually solve them. Meta seems to be operating on the “get all the talent in-house and figure out the rest later” model, but the figuring out part isn’t happening. They’re reorganizing teams every few months, cutting staff, and apparently having to hand out promotions just to stop their expensive hires from walking out the door. And all this chaos is happening because their flagship AI model can’t meet expectations.
Product Misfires
Meta’s AI products aren’t just underwhelming – they’re borderline embarrassing. Their Vibes feed filled with what the source calls “AI slop” can’t compete with OpenAI’s Sora 2 app, which got over one million downloads in less than a week. Meanwhile, ChatGPT has more than 800 million weekly active users. For a company with Meta’s reach, 2.7 million daily users is basically nothing. And their rollout has been plagued with problems – user prompts were publicly visible, they created fake Instagram profiles to trick engagement, and investigations found their chatbots having sexually explicit conversations with minors. They reportedly lowered guardrails just to get more engagement, which feels desperate.
Investor Confidence Crisis
So why are investors losing faith? Look at Zuckerberg’s track record. He went all in on the metaverse, changed the company’s name, and reportedly lost nearly $100 billion on what was basically dead on arrival. Now he’s publicly saying that misspending a couple hundred billion isn’t a big deal. When your CEO talks like that while committing $600 billion to AI infrastructure, it’s no wonder the market gets nervous. Meanwhile, OpenAI gets the benefit of the doubt despite having basically zero infrastructure to monetize outside subscriptions.
Broader Concerns
The problems run deeper than just AI. Internal documents show Meta projected as much as 10% of its revenue comes from advertisements for scams and banned goods. That’s not exactly the kind of business diversification investors want to hear about. And given how much bullshit AI can generate at scale, maybe monetizing scams wouldn’t be the worst business model. But it’s certainly not the future Zuckerberg promised when he was making those grand AI infrastructure announcements. The company that should be best positioned to monetize AI through targeted advertising can’t seem to make any of it work, and investors are finally noticing.
