Really appreciated this framing of AI as a hire you have to manage, not a magic box you occasionally query. The “context before tasks” and “iterate, don’t restart” sections especially mirror what separates high‑leverage founders and operators from everyone else.
Yes, the shift really is managerial. Most people treat AI like a search bar when it behaves far more like a junior colleague who needs context and feedback. The data supports that. A 2024 McKinsey study found only about 21% of organisations using generative AI have actually redesigned workflows around it. Most are still using it as a simple assistant.
The “context before tasks” point is the real unlock. Once the system understands standards and tone, the output improves dramatically.
The individual management frame makes sense - context before tasks, iterate don't restart. Where I've seen this break down is at team level. When everyone on a team is managing AI individually but nobody is managing what AI is doing to the team's collective reasoning, you end up with faster output and thinner understanding.
I wrote about this recently - the developers who can't debug without Copilot aren't lazy, they're just missing the layer of management above the tool.
Excellent article. It's more than just a tool, but a fundamental shift in who you are and what you do. We're all managers now, if we have the guts for it.
Thank you for this opportunity Daniel! Such an honor to contribute to your newsletter!
Managing tools well often matters more than the tools themselves.
This is especially true with AI nowadays.
The founders winning with AI are the ones who treat it like a real teammate. Context, feedback, systems. Same as managing people.
Really appreciated this framing of AI as a hire you have to manage, not a magic box you occasionally query. The “context before tasks” and “iterate, don’t restart” sections especially mirror what separates high‑leverage founders and operators from everyone else.
Agreed, Alex, that was one of my main takeaways too.
Yes, the shift really is managerial. Most people treat AI like a search bar when it behaves far more like a junior colleague who needs context and feedback. The data supports that. A 2024 McKinsey study found only about 21% of organisations using generative AI have actually redesigned workflows around it. Most are still using it as a simple assistant.
The “context before tasks” point is the real unlock. Once the system understands standards and tone, the output improves dramatically.
The individual management frame makes sense - context before tasks, iterate don't restart. Where I've seen this break down is at team level. When everyone on a team is managing AI individually but nobody is managing what AI is doing to the team's collective reasoning, you end up with faster output and thinner understanding.
I wrote about this recently - the developers who can't debug without Copilot aren't lazy, they're just missing the layer of management above the tool.
The magic quadrant on the x&y - ability to master people x ability to master agents - both 📈
AI is more than a tool; it changes the way we manage our businesses.
Excellent article. It's more than just a tool, but a fundamental shift in who you are and what you do. We're all managers now, if we have the guts for it.