UpTrajectory Review

Meta is quietly asking some individual contributors in its Applied AI division if they want to become managers again, reversing a piece of its aggressive flattening campaign. The company had reassigned roughly 7,000 employees to this new AI training group earlier this year, stripping many former managers of their titles and authority in the process. Now those same people are being invited back into management roles through an opt-in program. The move comes after Meta spent 2024 eliminating about 10 percent of its workforce with layoffs that disproportionately targeted middle managers, and after CEO Mark Zuckerberg spent months praising lean, startup-style organizational structures as the future of Big Tech.

For small-business operators watching Silicon Valley's management trends, this reversal carries a warning that flatter is not always better. Meta's flattening was never just about efficiency—it was about redirecting billions toward AI infrastructure while maintaining investor confidence. Small businesses lack that luxury of burning cash for strategic positioning. When you eliminate management layers, you also eliminate the people who translate between leadership vision and daily execution, who catch problems before they cascade, and who develop junior talent. Meta can afford to experiment because it has 7,000 people to reshuffle; a 20-person company that strips out its two team leads may discover too late that those roles existed for reasons beyond bureaucratic bloat.

What makes this genuinely notable is not the reversal itself but what it reveals about the limits of AI-driven efficiency promises. Zuckerberg's flattening rhetoric was explicitly tied to AI taking on more managerial and administrative work. Yet here is Meta's newest, most AI-centric division finding that it still needs human managers. The source also notes that Meta already had to let some AAI employees leave the unit after internal backlash over being 'drafted' against their will—suggesting organizational strain that management layers might help absorb. Meanwhile Uber, another flattening enthusiast, just cut half its 'micro-teams.' The fad is faltering because the theory ignored how coordination actually works at scale.

The downstream effects split unevenly. For Meta employees, this creates a fractured career landscape: some former managers get a path back while others remain individual contributors, likely breeding resentment and confusion about what signals advancement. For the broader tech industry, it means the 'manager-lite' playbook that venture capitalists and consultants were pushing may lose its fashionable status, which could actually help startups that resisted the trend. For Meta's competitors, there is recruiting opportunity in poaching disillusioned talent who experienced two reorgs in under a year. And for anyone selling organizational consulting, expect a pivot from 'radical flattening' to 'intelligent hierarchy' with fresh PowerPoint decks.

Watch whether Meta extends this manager restoration beyond AAI or keeps it contained to its most technically complex division. If it spreads, the flattening era ends with a whimper, not a bang. Also watch how Meta handles the optics: Zuckerberg has tied his personal leadership brand to lean efficiency, so admitting some management layers were necessary after all requires narrative gymnastics. For operators reading this, the actionable insight is to audit your own management structure against function rather than fashion. Count spans of control, measure where decisions bottleneck, and ask your team leads what they actually do all day before eliminating their roles. Meta's billions bought it the right to learn this lesson publicly. You will not have that budget.

Takeaway: Audit management roles against actual function, not Silicon Valley fashion—Meta's reversal proves flattening has hard limits.