The Rise of the AI Player-Coach Leader
On May 5th, Coinbase announced layoffs of 14% of their workforce in yet another let’s-blame-AI announcement. “Adapting for the AI era”. The push for AI-driven workflows and higher productivity per employee.
But this announcement held something new: a clear, if stark, message for leaders.
“- No pure managers: Every leader at Coinbase must also be a strong and active individual contributor. Managers should be like player-coaches, getting their hands dirty alongside their teams.”
— Coinbase layoff announcement (delivered on x.com, because of course it was.)
That term - “player-coach” is one I’ve been circling for over a year. It started when I was talking to founders or CPOs/Heads of Product at small, agile companies in the Bay area about how they’re working. They were actually pushing code or fixing bugs themselves for the first time in decades. They were working with their CTOs to get their product team access to code bases or pipelines to push features to production, at least behind A/B tests. They were doggedly pushing how far AI CAN take you if you trust it - even when you shouldn’t - and purposely seeing where it breaks, then patching with guardrails and moving on until it breaks again. The goal: to move as fast as you’ll be able to some day, when reliability catches up.
And it IS catching up.
If you’ve had at least a passing glance at tech twitter, you’ll know that programming juggernauts like Andrej Karpathy (co-founder, OpenAI) have been screaming from the rooftops that software development has completely changed since December 2025.
“It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the “progress as usual” way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since[…]imo, this is nowhere near “business as usual” time in software.”
— Andrej Karpathy on x.com, Feb 25 2026
The world’s top developers, the most skeptical, attention-to-detail pedants on the planet, are trusting LLMs for long-running and complex tasks. And if you’re not asking what that means for product and business leaders in the software development ecosystem - you should be.
The danger is waiting to understand the difference between the ‘pure manager’ and the ‘player-coach’ until it’s too late. And the answer is not simply loading up on tech skills to become the newly-vaunted ‘super-IC’, or to take on two jobs and 15+ reports until we burn out. There is another way. I know, because I’ve seen the leaders quietly making this shift: layering AI fluency and practice on top of the strategic judgment, organizational instinct, and discernment they’ve already built over their careers. It’s powerful, and it’s what is making them indispensable at their organizations right now - even amid the layoffs and chaos.
You’re not starting over. You’re adding depth. That’s a different path, and one I wanted to draw back the curtain on. It’s not a path that needs programming experience - it’s one that requires curiosity, persistence and an excitement to actually PLAY. Not to be the star player - but to show your team what it means to really love the game.
If you’re a Director+ product leader eager to see what ‘getting on the field’ looks like - from the skills needed at the leadership level to stories from player-coaches in the trenches - welcome to The AI Player Coach.
More soon. Pass it on.