Meta's AI Ambitions: New Crew, Same Old Tricks

Meta, the social media juggernaut formerly known as Facebook, is assembling what it hopes will be an elite force of AI researchers. The mission? To keep us endlessly scrolling on Facebook and Instagram. Naturally.

Meet the New AI Team

Heading this elite squad is Yang Song, a former TikTok exec who's now Meta's vice president of recommendation research. Song jumped ship in November 2025, presumably with dreams of revolutionizing Meta's recommendation systems. But is this just another layer of the company's already formidable advertising apparatus?

The new unit, MRS Research, has been quietly hiring stars from AI powerhouses like OpenAI, Google, and Amazon. It's become a talent tug-of-war, but will this influx of intellect really change Meta's game, or is it just more of the same ad-driven strategy?

The Talent Carousel

Lihong Li from Amazon and Xiaolong Wang, previously of OpenAI, are some of the latest recruits. Add to that Google researcher Fei Sha, and you've a team that reads like a who's who of AI experts. But let's not kid ourselves, folks. The press release said innovation. The 10-K said losses.

Meta's been busy reshuffling since summer 2025, even creating the Meta Superintelligence Labs under Alexandr Wang, formerly of Scale AI. Yet MRS isn't part of this new lab. Confused? So am I. But maybe that's the point.

Can AI Save Meta?

Meta's long had its sights set on optimizing the algorithms that feed its ad revenue, a strategy that seems to be as old as time. In late 2025, they launched an AI model promising better ad performance. But spare me the roadmap. What Meta really needs is a strategy that doesn't just lean on AI like a crutch. Or is AI just a shiny distraction from deeper issues?

Let's be honest: AI talent might be Meta's new shiny toy, but if the underlying business remains unchanged, are we really expecting different results? I've seen enough. The optics suggest a high-stakes game of AI one-upmanship, but without a tangible shift in strategy, it's all noise.