I keep hearing from founders that they feel behind, like they don’t have the time, and like they wish they could spend a week or two just going really deep on AI.

But the problem is that by the time you finish “catching up,” the tools will have changed again…

The thing is, you don’t catch up on AI by reading about it.

You catch up by using it for everything that you do.

Here’s what I recommend: Use it as your doorway into work.

Start every task AI-first and see where it takes you. This allows you to learn the boundaries. Getting frustrated because it won’t do what you need to is a sign that you’ve hit a current edge.

You start to see what it can do, where it breaks, what still needs your judgment, and which parts of the work you actually want to keep doing yourself.

One example for me right now is design using Canva.

I’ve connected from Codex to Canva because Codex design still sucks. Even then, the connection between Canva and Codex isn’t perfect. So that type of design is still a bit behind.

So then I’m forced to test another tool, try another workflow, and sort of hack my own system. That’s how you build taste and understanding with these systems.

Then, once you’ve gained proficiency, I think you can follow a more human rule:

• Automate anything that can and should be automated.

• Don’t automate anything you enjoy doing.

Your business is becoming more like software. That means we’re taking daily operations and turning them into automations and flows that can be owned by individual agents.

To learn that language, start every task in an LLM and see how far it takes you.

xx David

Leave a comment

Share

I’ve AI-pilled a number of founders recently.

These are thoughtful, tech-forward people. They are not resisting the future. Mostly, they just feel like they haven’t had the time to catch up with what is now possible.

They keep telling me, “I just don’t have the time.”

Then, once I get them going on Codex or other platforms, their perception changes quickly. They’re still thinking about a few months ago with Openclaw and all of the setup time. Now they realize they can write, research, prototype, analyze, build, and operate with a level of breadth that used to require a much larger team.

Teams, though, have been harder to AI-pill. And founder’s are complaining about it. That is creating a strange effect inside companies. AI is supposed to make organizations more productive, but in the short term I think it may be consolidating more power at the top.

The founder gets faster first.The exec team gets more capable first.

And because the rest of the team is slow to adopt, more responsibility starts flowing back upward in the org.

That is probably a net negative, but it is also a natural response.

AI tools give strong operators more range and they reward people who can make decisions, test quickly, and own the messy middle between idea and execution.

Founders are usually hungry to learn this stuff because they can immediately feel the leverage. Most teams do not yet have the habits, permissions, or operating norms to use these tools well.

So, we are still missing the middle layer.

Plus there are all of these thorny problems we haven’t solved yet.

• How do you safely sandbox company data?

• Who gets read/write access to what?

• What should AI be allowed to touch?

• Who owns the decision when an AI-assisted workflow changes part of the business?

There are no standard practices yet. So for now, the organization starts to flatten, but not in the healthy way. The founder becomes more capable before the company becomes more capable.

I still don’t think the major productivity wave has fully hit small companies yet. I’m excited about this major tailwind thats’ going to hit startups with small teams who could use the boost.

Turns out we need to teach employees to think more like Founder’s.

That might take a while…

xx David

Share

Leave a comment