Feature
Research Report
The State of the Digital Workplace
Dacher Keltner has spent decades studying what power does to people. One of his more uncomfortable findings? The more power someone accumulates, the less accurate their read of other people becomes. They stop picking up social cues, and they start confusing their own perspective with reality.
The shift happens gradually, and it happens to almost everyone.
Most CEOs of large companies have been insulated from honest feedback for long enough that they've lost the ability to accurately assess what their employees actually experience. The communication programs designed to fix this often make it worse. They're built to make leadership feel connected rather than to actually connect anyone.
So the “listening” programs proliferate. You’ve probably seen them: Town halls, all-hands Q&As, internal social channels, CEO video series, skip-levels and advisory councils. Each one generates enough of the appearance of connection. None of them closes the gap.
At a certain scale, that gap becomes structural. Most companies haven't figured that out. And at least one recently decided the answer was to build something else entirely.
The Metaverse Failed. Workplace Failed. Now, Meet AI Zuckerberg.
On April 13, the Financial Times reported that Meta is developing a photorealistic AI version of Mark Zuckerberg, trained on his mannerisms, earnings call transcripts, public statements and internal writings.
The goal? Give the company's 79,000 employees something they can actually interact with. A simulation of the founder available at a scale the actual founder isn't.
Zuckerberg has reportedly been spending five to 10 hours a week writing code on AI projects and attending technical engineering reviews, an unusual level of hands-on engagement for someone running a $1.6 trillion company. The AI version of him is, in part, how the other thousands of people who work there are supposed to feel his presence.
For any other company, this may seem bizarre, but Meta has been here before.
In 2021, the company renamed itself around a vision of immersive virtual work. I covered my first meeting in the metaverse for Reworked, going in skeptical and finding something with genuine potential buried under enormous behavior change requirements that most organizations didn't have the patience for.
The change never came. Reality Labs has since lost over $80 billion, faced multiple rounds of cuts, and the metaverse pitch has gone dormant.
Workplace from Meta followed a similar arc. In May 2024, the company told its roughly 7 million paying customers it was shutting down the enterprise social network to focus on technologies that would "fundamentally reshape the way we work."
The assumptions didn't change between the metaverse and Workplace and AI Zuckerberg. Only the technology did.
The Difference Between Connection and the Performance of Connection
Google's weekly TGIF all-hands is the example that comes up whenever someone tries to argue that CEO connection at scale is possible. Larry Page and Sergey Brin took unfiltered questions from the entire company, live, on any topic.
And for a while, it worked.
What employees actually had wasn't access to the founders. It was accountability with an audience. They could watch leadership get asked things they hadn't prepared for and see what happened. That's a structural condition, not a communication program. When the questions started getting softer as Google grew and the answers got more polished, employees felt the difference. Eventually, the company stepped back from the format altogether.
Most companies never build the real version to begin with.
Researchers at Carnegie Mellon and Emory surveyed workers and managers about AI systems designed to stand in for supervisors. Workers worried about job displacement, accountability gaps and surveillance.
In short, they wanted human relationships.
One subset pushed back on the premise entirely. Some Gen Z workers thought an AI boss might actually be fairer and less intimidating than a human one. That finding says less about AI than it does about how much some employees distrust their managers.
What the research points toward consistently is that connection at scale gets built close to the ground, whether that’s through managers who are genuinely invested, feedback that demonstrably changes something or peer relationships that develop horizontally.
None of it requires the founder. All of it requires sustained investment in the layers between the top of the org and the people doing the work.
Conference
Apr
22
HR Tech Europe Amsterdam 2026
Register
Conference
Apr
27
Gartner Digital Workplace Summit London 2026
Register
Webinar
May
6
Retention Is a Communication Problem: New Data from 6,000+ Employees
This session gives leaders a clear diagnosis and practical path to treat retention as an infrastructure problem.
Register
Conference
Jun
2
Gartner Application Innovation & Business Solutions Summit Las Vegas 2026
Register
Webinar
On demand
Do Learning Programs Really Work? How to Turn Education Into Engagement In Healthcare
See how leaders are using learning programs to build trust with healthcare professionals and create measurable engagement.
Watch Now
Webinar
On demand
Fuel Your Front Line: Workforce Trends Shaping 2026
Get expert insight into the tools and tactics top frontline teams will need in 2026.
Watch Now
What Three Years of Layoffs Do to a Trust Account
Understanding why AI Zuckerberg exists requires understanding what the last few years have felt like from inside the company.
Since 2022, the company has cut roughly 25,000 jobs, each round with its own framing. The January 2025 cuts were described publicly as targeting low performers, a characterization that turned out not to match the performance reviews many of those employees had received.
The return-to-office experience added its own complexities. Meta requires three-plus days a week in the office with daily attendance tracking. Employees have described coming in to spend the day on Zoom with teammates in other offices, unable to book conference rooms for in-person work. The company that spent years selling virtual work as the future couldn't make its own hybrid policy function.
And now we have a conversational AI trained on Zuckerberg's public statements, designed to help employees feel more connected to him.
A more articulate version of his public positions, available on demand, doesn't close the distance between leadership and the workforce. It's a simple reminder about the vast gulf between an out-of-touch CEO and his employees.
Connection Is an Organizational Problem. Meta Treated It Like an Engineering One.
The Carnegie Mellon-Emory research shows workers would accept AI for narrow, transactional uses, like relaying information, handling routine questions and cutting administrative friction. That's a real and relatively uncontroversial application.
What Meta built reaches for something different: the felt experience of being known by the person at the top.
That was never a technology problem to solve. The hard answer requires investment in the unglamorous organizational work that produces nothing you can announce to the tech press. Moves like better managers, real feedback loops and advisory structures that change decisions.
You know. The things that employees have repeatedly said they cared about.
That work is hard and slow. It’s tech-assisted, not a tech solution.
Which is probably why Meta built something instead and will probably build something else in the future. Still, it’s hard to say AI Zuck is worse than legless metaverse Zuck, waiting for you dead-eyed in a fake conference room.
Editor's Note: What if it turns out all of this was a performance art piece?
- Employees Crave Connection: Here's How to Build It — Connection at work can be both emotional and intellectual. Understanding what kind of connection you're trying to build is the first step.
- Enterprise Social Isn't Over, It's Just Changing — Ultimately, enterprise social networks aren’t dying; they’re evolving. Their success hinges on culture, not technology.
- Companies Are Not Families — Sometimes it's hard to reconcile actions with words.