I’m going to say something that might get me kicked out of a few Silicon Valley dinner parties: the concept of a centralized “AI Coworking” space is mostly a fantasy. In fact, chasing that fantasy cost one of my portfolio companies over $100,000 in lost productivity and subscription fees before we pulled the plug. The promise was a seamless, AI-driven utopia for our remote team. The reality was a chaotic mess of missed context and frustrated engineers.
When we first signed up for a top-of-the-line AI coworking platform at RemoteTeam, I was all in. The pitch was irresistible. Imagine a virtual office where AI not only summarizes every meeting but also anticipates your information needs, drafts your reports, and even helps mediate discussions. It was supposed to be the ultimate tool for asynchronous work, a digital headquarters that would make us more connected and efficient than ever. For the first two weeks, it felt like magic. The AI summarized a sales call I missed, and it was pretty accurate. It felt like we were living in the future.
The Cracks Begin to Show
The honeymoon ended quickly. The first major problem was the noise. The platform was designed to be the single source of truth, a central hub for everything. But instead of creating clarity, it created a firehose of notifications, AI-generated summaries, and action items. The constant pings and updates made deep work impossible. We were spending more time managing the AI than doing our actual jobs.
I remember one critical moment. We were pushing a major update. A junior developer flagged a potential security vulnerability in a side channel. The AI, in its infinite wisdom, decided this was a low-priority comment and buried it in its daily summary. The lead engineer never saw it. We caught it in the final QA check, purely by luck. That near-miss cost us a week of delays and a lot of stress. The AI that was supposed to connect the dots had erased the most important one. It was a sobering lesson: nuance is everything, and the AI was deaf to it.
The Performance Review Disaster
The biggest failure, by far, was when we tried to use the platform’s AI for performance reviews. On paper, it was a dream for any founder who champions remote work. The AI would analyze contributions across all our systems—code commits, document edits, messages—and provide an “objective” performance score. No more human bias, just pure data.
It was a catastrophe. The AI gave our best engineer, a quiet but brilliant problem-solver who spent his time mentoring others and tackling our gnarliest architectural challenges, a mediocre rating. Why? Because he didn't write a lot of comments or participate in the noisy channels. His real contribution wasn't easily quantifiable. Meanwhile, a much less effective employee who was great at talking a big game scored in the top 10%. The AI mistook activity for achievement.
I had to personally step in and do damage control. It took a two-hour call to walk our top engineer back from the ledge. He felt completely devalued, and frankly, he was right. We had outsourced our judgment to an algorithm that couldn't tell the difference between a foundational piece of code and a flurry of trivial messages. We switched the feature off that day and never looked back.
The Turnaround: From AI Coworker to AI Assistant
I was ready to write off the whole experiment. But we knew AI had to be part of the answer. We just had the wrong approach. The problem wasn't the AI itself; it was the “coworking” paradigm. We were trying to force AI to be a member of the team, a central brain. It failed miserably.
So we made one fundamental change that shifted everything. We stopped trying to build an AI-powered team and instead focused on empowering each team member with their own personal AI assistants.
Instead of one monolithic platform, we adopted a stack of specialized, best-in-class AI tools that individuals could control. We didn't need an AI to watch over us; we needed tools that worked for us.
Here’s what that looked like in practice:
Meetings: Instead of a central AI summarizing every call for everyone, each person uses their own transcription and summarization tool. They can ask it specific questions relevant to their work, like, "What were the final specs agreed on for the new API?" This is far more effective than a generic summary pushed to ten people who don't need it.
Coding: Our developers use tools like GitHub Copilot. The AI helps them write better code, faster. It’s a pair programmer that never gets tired, not a manager looking over their shoulder. It augments their skill, it doesn't try to measure it.
Writing and Communication: We use AI to help with first drafts, to clean up grammar, and to translate. This is a huge boost for a global team. The key is that a human is always in the loop to add the final context, tone, and strategic thinking.
Async-First: We reinforced our async communication principles. The goal is not to have an AI participate in the conversation, but to use AI to make our own contributions more thoughtful and clear. An AI can help you structure a proposal or summarize a long thread before you jump in, so your input is more valuable.
The result? Our team's measured efficiency, based on project completion velocity, jumped by over 50%. More importantly, morale skyrocketed. People felt trusted and empowered, not monitored. They had powerful tools at their disposal that they could use as they saw fit.
My Contrarian Take
The future of AI in the workplace isn't a digital panopticon or a virtual office with an AI hall monitor. It's a distributed network of specialized, personal assistants that make each of us better at our jobs. The goal of AI should be to augment human capability, not to replace human judgment or connection.
So my advice to other founders is this: stop chasing the AI coworking dream. It’s a solution in search of a problem. Instead, ask your team what their biggest bottlenecks are and find targeted AI tools that solve those specific problems. Give your people the power and the trust to use them. That’s how you build a real team of the future.
Frequently Asked Questions
How has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.