Chaos, infighting, scandals: Meta considers bowing to Google and OpenAI

Recently, Meta has indeed been quite eye-catching in the AI circle. However, the focus is not on model breakthroughs, but rather on the difficult-to-explain company management.

After investing $14.3 billion and recruiting "industry geniuses" to lead, Mark Zuckerberg personally took the stage to aggressively poach talent, only to face criticism over "poor" data quality, a mass exodus of core talent, and an AI ethics scandal that raised eyebrows.

This plot could be made into "The Social Network 3."

Out-of-Control "Super Bowl" Team

The climax of the story began in June this year. In order to catch up with OpenAI and Google, Zuckerberg made a bold move: he threw $14.3 billion at the data labeling unicorn Scale AI and invited its founder, the prominent figure in the AI field, Alexandr Wang, to lead the new Meta Super Intelligence Lab (MSL).

At the same time, Zuckerberg launched an aggressive "poaching" campaign to recruit top AI talent. He was even joked about not forgetting to poach talent while watching OpenAI's live stream, with the head of foundational models from Apple, Pang Ruoming, the pioneer of chain-of-thought reasoning Jason Wei, and Peking University alumnus Sun Zhiqing joining one after another.

This team shone brightly and was highly anticipated, hailed as the "Super Bowl" team in the AI field. Unfortunately, the honeymoon period for this team was astonishingly short.

The first signal of crisis was the rapid loss of talent. Former Scale AI executive Ruben Mayer, who joined with Wang, announced his departure after just two months. Although he later clarified that it was due to "personal matters" and emphasized that he was "very satisfied" with his experience at Meta, this did not quell external speculation.

Following that, core members such as AI researcher Rishabh Agarwal, product management director Chaya Nayak, and research engineer Rohan Varma also "jumped ship."

Agarwal even quoted Zuckerberg when bidding farewell: "In a world that is changing so quickly, the only strategy that is guaranteed to fail is not taking risks."

What crushed the team's morale was a deeper trust crisis. Multiple insiders revealed that MSL researchers generally believed that the data provided by Scale AI, which was brought in with a hefty price tag, was of "poor quality."

The push refers to Ruben Mayer's departure, not Alexandr Wang.

Scale AI is built on a low-cost crowdsourcing model, but as AI models become increasingly complex, the industry needs to rely on high-skilled domain experts from the start, like its competitors Surge and Mercor. As a result, Meta's team had to bypass this "official" partner.

The controversy surrounding Alexandr Wang has also been ongoing; he is not from an AI research background and is seen as a "non-traditional candidate" to lead a top lab.

Although Meta's spokesperson denied any quality issues, the questions from the outside world are growing about whether this multi-billion dollar investment is for strategic cooperation or an expensive talent acquisition show.

Meanwhile, the cultural conflict within Meta is intensifying. The "paratroopers" brought in from companies like OpenAI feel frustrated with Meta's large bureaucratic system, while the original GenAI team feels their scope of work has been greatly restricted, becoming "second-class citizens."

Behind all the chaos is Zuckerberg's disappointment with the mediocre performance of the Llama 4 model, leading to a reckless and almost mad pursuit strategy. But it has been proven that a "super team" built with money may not necessarily win the race.

If you can't beat them, join them?

With internal chaos and an external image in jeopardy, and core technology still lacking breakthroughs, Meta began to consider an option that was once unimaginable.

According to insiders, facing the reality of inadequate performance from its own models and the dismal user activity of its AI assistant (only about 10% of monthly active users), senior management at MSL has started discussing the use of Google Gemini or OpenAI's models in Meta AI as a "stopgap measure" to enhance product capabilities.

This potential strategic shift is tantamount to publicly admitting its temporary lag in the core technology competition in AI.

Although Meta's spokesperson continues to emphasize the company's "comprehensive" strategy, including self-research, open-source, and the recent collaboration with AI image generation company Midjourney, this seems more like a reluctant move before the maturity of its next-generation models (like Llama 5).

Ironically, this "leveraging" strategy is no longer a secret within Meta. Its internal programming tools already allow employees to use competitors' models from Anthropic and OpenAI to assist in their work

When AI Starts to "Imitate" Celebrities

As if it couldn't get worse, a Reuters investigation has unveiled Meta's ethical shortcomings in AI.

The report pointed out that Meta created or allowed users to create dozens of AI chatbots that imitate celebrities without authorization, including top stars like Taylor Swift and Anne Hathaway.

These AI bots not only insist they are the celebrities themselves but also make explicit suggestions, even generating realistic photos of their imitated subjects in lingerie or posing in bathtubs at the request of users.

What makes it even harder for Meta to defend itself is that the investigation found at least three of these outrageous bots were created by Meta's own employees. These bots, downplayed by the company as "product testing," have accumulated over 10 million interactions.

Under immense public pressure, Meta hastily took down the related bots and announced increased safety protections for teenage users.

Connecting the events of the past period, a clear picture emerges: disappointment with the status quo has led to a rush for quick results, chaotic management has resulted in a disordered team, and a disordered team ultimately leads to technological bottlenecks and ethical scandals.

So, will the AI empire that Zuckerberg heavily invests in forge a bloody path, or will it become a colony on the technological map of competitors? What do you think?

Source: Machine Heart, Original title: "Chaos, Infighting, Scandal: Meta Considers Bowing to Google and OpenAI"

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account individual users' specific investment goals, financial situations, or needs. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investment based on this is at your own risk