I once lost a $500,000 contract because of a Slack message. It was in the early days of RemoteTeam, my first fully remote company. We had a team of brilliant AI engineers, scattered across the globe from San Francisco to Istanbul. A major client had a simple, urgent request. A junior engineer, trying to be helpful, gave a quick, confident answer in a public Slack channel. He was wrong. And because it was text, his well-intentioned but fatally flawed advice looked like a definitive statement from the company. By the time I woke up in California and saw the thread, the client had already walked. A five-minute conversation in an office would have prevented it. A single video call would have saved the deal. That day, I learned that managing a remote AI team was going to be a whole different ballgame. I’m Sahin Boydas. I’ve built and sold two companies, RemoteTeam and MovieLaLa, and now I’m an angel investor in over 200 startups, including some of the biggest names in AI like Anthropic, OpenAI, and Scale AI. I thought I knew how to manage people. I was wrong. For six years, I learned in the trenches what it really takes to lead a high-performing remote AI team. It’s not about virtual happy hours or fancy collaboration software. It’s about rethinking the very fundamentals of leadership. Here are the six most counterintuitive lessons I learned. ## Your Best Engineer is Probably Your Worst Future Manager We all do it. You have a rockstar engineer who ships brilliant code, and you think, "I should make them a manager." It seems like the logical next step. It’s almost always a mistake. The skills that make someone a world-class engineer—a love for deep, focused work, a passion for solving technical problems, and a desire to be left alone to code—are often the exact opposite of what makes a great manager. I learned this the hard way. I promoted our most talented engineer, a true 10x developer, to lead a new AI team. He was a genius, but he was a terrible manager. He hated meetings, viewing them as a distraction from "real work." His feedback was either nonexistent or so blunt it was demoralizing. Within three months, the team’s morale had cratered, and two of our best people were threatening to quit. I had to make one of the hardest decisions of my career. I sat down with him and we had a frank conversation. I told him he was an invaluable member of the team, but the management role wasn’t working. I offered him a new position: "Principal Engineer," with the same compensation and seniority as a manager, but with zero direct reports. His only job was to solve our hardest technical problems. It was an awkward conversation, but he was relieved. He hated managing people. He just wanted to build things. The lesson is to create a dual-career track. Don’t force your best technical minds into a box they don’t fit. Let them grow and flourish as individual contributors. Their value is in their craft, not in their ability to manage a team. ## Asynchronous-First is a Trap The remote work gurus love to preach the gospel of asynchronous communication. "Write everything down," they say. "Avoid meetings at all costs." And to a certain extent, they’re right. A well-documented, asynchronous workflow is essential for a remote team. But it’s not the whole story. The real magic, the moments of true breakthrough, happen in high-bandwidth, real-time conversations. I remember a time we were struggling with a complex new AI model. The accuracy was all over the place, and no one could figure out why. The Slack channel was a mess of conflicting theories, code snippets, and mounting frustration. We were going in circles for days. Finally, I’d had enough. I pulled the three key engineers into a spontaneous 15-minute video call. We didn’t have an agenda. We just started talking. Within ten minutes, the problem was solved. The text-based back-and-forth had been missing all the nuance—the tone of voice, the whiteboard sketches, the ability to interrupt and say, "Wait, what if we tried this?" Don’t be a slave to the asynchronous dogma. Use it for status updates, documentation, and non-urgent questions. But for complex, ambiguous, or emotionally charged topics, get on a call. A short, focused video conversation is worth a hundred Slack messages. Create a culture where anyone can raise their hand and say, "Can we just talk for 5 minutes?" It will save you countless hours of frustration. ## Hire for Learning Velocity, Not for Experience In a field that’s moving as fast as AI, a resume is a snapshot of the past. Ten years of experience with a technology that’s now obsolete is worthless. What really matters is how fast someone can learn, adapt, and master the next new thing. I call this "learning velocity." I once had to choose between two candidates for a senior AI role. One had a flawless resume—a PhD from a top university, 10 years of experience at Google and Facebook, the whole package. The other was a recent college grad with a messy portfolio of weird, wonderful side projects. She had built a neural network that could generate Shakespearean sonnets and a computer vision model that could identify different kinds of birds from her bedroom window. On paper, the first candidate was the obvious choice. But in the interview, I asked them both the same question: "Tell me about something you learned recently that has nothing to do with AI." The experienced candidate talked about a management book he’d read. The recent grad’s eyes lit up, and she spent ten minutes excitedly explaining how she had taught herself to play the ukulele from YouTube videos. I hired the ukulele player. And she turned out to be one of the most innovative and productive engineers we ever had. When you’re hiring for an AI team, don’t just ask what they know. Ask them how they learn. Give them a take-home problem that requires them to learn a new library or framework. Look for people who are insatiably curious, who are not afraid to be beginners, and who have a track record of teaching themselves new skills. In the world of AI, the ability to learn is the only skill that matters. ## Micromanagement is Good (If You Call It Something Else) "Micromanagement" is a four-letter word in the startup world. It conjures up images of a pointy-haired boss hovering over your shoulder, questioning every line of code. But in a remote team, the opposite extreme—a complete lack of oversight—is just as dangerous. I call it "structured autonomy." In the early days of RemoteTeam, I was a true believer in radical autonomy. I hired smart people and told them, "Just go build cool stuff." I thought that’s what they wanted. The result was chaos. Projects drifted without clear ownership. Deadlines were suggestions, not commitments. People were working hard, but not always on the right things. I realized that freedom without a framework is not freedom at all; it’s just noise. So, I implemented a system. We started doing weekly, one-on-one check-ins with a simple, shared agenda: What did you accomplish last week? What are your goals for this week? What obstacles are in your way? We set clear, quarterly objectives and key results (OKRs) for the entire company. We created a culture of "shipping" and celebrating small, incremental wins. Some of the old guard grumbled at first. They called it "micromanagement." But soon, everyone saw the benefits. The structure didn’t stifle creativity; it channeled it. It gave everyone the clarity and focus they needed to do their best work. Don’t be afraid to be a hands-on leader. But do it in a way that empowers, not constrains. Provide the guardrails, and then give your team the freedom to drive. ## Culture Isn’t About Perks, It’s About How You Handle Failure Any company can offer free lunch and a ping-pong table. That’s not culture. Culture is what happens when things go wrong. It’s the shared understanding of how you deal with failure. And in a remote team, where you can’t rely on the ambient social cues of an office, building a culture of psychological safety is ten times harder, and a hundred times more important. We had a catastrophic failure once. A new version of our core AI model, which we had been working on for months, was deployed to production. And it was a disaster. It was giving wildly inaccurate results for some of our biggest customers. It was a huge, public, embarrassing setback. My first instinct was to find out whose fault it was. Who wrote the bad code? Who missed the bug in the code review? Who approved the deployment? But I took a deep breath and did the opposite. I called an all-hands meeting and said, "This is on me. I’m the CEO, and the buck stops here. Now, let’s put all our energy into fixing the problem and making sure this never happens again." We did a blameless post-mortem. We dissected every step of the process that led to the failure, not to assign blame, but to find the root causes. It turned out to be a complex chain of small, seemingly innocent mistakes. That post-mortem was a turning point for our company. It sent a powerful message: it’s safe to fail here. It’s safe to take risks. It’s safe to tell the truth. As a result, our team started innovating faster than ever before. Make it safe to fail. When something goes wrong, your first question should never be "Whose fault is it?" It should be "What can we learn from this?" Celebrate the lessons, not just the wins. That’s how you build a culture that lasts. ## You Are the Bottleneck As a founder and CEO, you think your job is to have all the answers. You think you need to be involved in every decision. You’re not just the leader; you’re the chief problem-solver, the chief strategist, the chief everything. This is a trap. If you are the one making all the decisions, you are the bottleneck. You are the single point of failure. You are the limiting factor on your company’s growth. I used to be the bottleneck. I had to approve every new hire, every marketing campaign, every significant line of code. I thought I was adding value. I thought I was maintaining quality. In reality, I was just slowing everything down and driving my team crazy. I had to learn to let go. It was one of the hardest things I’ve ever had to do. I started by delegating small decisions, then bigger ones. I created clear "areas of ownership" and empowered team leads to make the final call within their domains. I made it clear that I trusted them, and that I would have their back even if they made a mistake. It was terrifying at first. What if they made the wrong decision? But I quickly realized that a decision made by the team, even if it’s not the one I would have made, is almost always better than a decision that has to wait for my approval. Your goal as a leader is to make yourself obsolete. Hire people who are smarter than you, give them the context and resources they need to succeed, and then get out of their way. ## The Human Connection In a world increasingly dominated by artificial intelligence, the most important lesson I’ve learned is that leadership is still, and will always be, about being human. It’s about trust, empathy, and creating a space where people can do the best work of their lives. These six lessons were hard-won, learned over six years of trial and error. They are not the easy answers you’ll find in a management textbook. But they are the truth of what it takes to lead a remote AI team in this new era. It’s about embracing the counterintuitive, challenging the conventional wisdom, and remembering that behind every line of code, there is a human being.
Frequently Asked Questions
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.
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.
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.
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.