Nobody Can Learn AI for You

Today I watched David Senra's interview with Tobi Lütke, Shopify's CEO. What caught my attention is that he views AI less as a corporate initiative and more as a practical, default capability that every executive across the organization should master. Most CEOs running a company that size don't talk about it that way.

My last post covered Bloomberg's piece on the Chief of Staff role — a job being recruited and amplified across the board to help CEOs adopt AI in the C-suite, among other functions. Lütke goes the other way, and so do I. The responsibility to learn these systems and extract the most from them sits with the individual.

In an April 2025 memo to his organization, which he later posted publicly, Lütke wrote that "reflexive AI usage is now a baseline expectation at Shopify." Employees were expected to instinctively test whether AI could help their core functions before approaching their work through traditional processes.

That is exactly the approach I adopted for myself in Q4 2022. The moment the first LLMs became available to the general public, I knew this was taking us back to when personal computers began to proliferate in homes when I was younger. Back then, I knew this equipment, with something called the internet via dial-up, opened a world of untold possibilities. Fast-forward 30 years, and I see the same renaissance once more, and just like back in the 80s and 90s, to me this is a matter of adopting its use or being left behind. Period.

In many organizations, AI is left to the Chief Technology Officer or a similar role. Lütke turned that rhetoric into operating rules that affect how work, staffing, and individual performance are evaluated. His approach is what I believe every company, no matter its size, should adopt. From the memo:

1. It starts at the top. "This applies to all of us—including me and the executive team."
2. AI should be the habitual first choice in daily work, not an occasional tool for approved projects.
3. Teams use AI before defining or building a project, while they are still trying out ideas, learning what is possible, and deciding whether a concept is worth pursuing. "The prototype phase of any GSD project should be dominated by AI exploration."
4. "Before asking for more Headcount and resources, teams must demonstrate why they cannot get what they want done using AI."
5. AI usage questions go into the performance and peer review questionnaire.

Why does it matter? Lütke is doing more than betting AI will raise productivity. He makes AI fluency a condition of standing. A team's case for hiring, for funding, or for keeping the way it already works now begins with a question he puts plainly: "What would this area look like if autonomous AI agents were already part of the team?"

This is highly practical. Whether you operate a multi-million-dollar company with hundreds of employees or a small family-run organization, you can apply these principles today. Everyone else is still catching up.