Most people are still treating AI like a software application or platform. That might be holding you back.

Last week, we looked at how differently people accept change. This email is about what to do once they’re willing to change.

I had a conversation this month with someone working hard to get more out of AI for himself and for his business. He's a smart guy, and he’s making progress quickly.

But he was approaching these tools the way we all learned to approach software. You learn what it is capable of, you learn how to use the interface, and then you start entering data or dragging and dropping to get work done.

That kind of approach was limiting the results he got back.

The shift that changes everything

Don't think of AI as just the newest type of software. Try this instead:

Think of your AI tools as able to completely do the work for you - not just give you answers to a search or request.

Think of your AI tools as digital employees.

Imagine you've just hired a new employee with an advanced degree. Enormous knowledge, real capability. But absolutely no idea how your business works.

You would not hand that person the keys on day one. You would onboard them, show them the ropes, give them guidelines and context for how your organization works and who does what.

The same three questions apply to these new digital employees:

• How much (or little) responsibility should I give this new employee right now?
• What information, data and training do they need access to?
• What context do they need about our business and how we actually operate?

Start small, then extend the leash

The principle of gradual increasing responsibility works whether you're using something like Claude Cowork where you work with it to accomplish a single task, or if you are using a more advance agent approach where you give it a job to do on a regular basis. (I've been using a Hermes agent for a few months, and just now starting to experiment with GrokBot. Really cool capabilities in these new digital employees.)

Give your digital employee a task with a tight scope and low risk. Watch the work and check it. Correct it, and give the digital employee your feedback. Then rinse and repeat.

It's the same approach you'd take with a human employee, right?

As your new digital employees proves it can do the job safely and the way you want it done, you extend more responsibility.

That is not a technology process. That is Management 101.

It’s the same thing good managers have always done with their people.

Which is the real point here

Management skills are about to matter a whole lot more in the future, not less, in my opinion.

The leaders who thrive over the next few years will be the ones who can run a team that includes both humans and digital employees. They will know the difference in how each one needs to be led and managed.

Most of your people were never formally taught to manage anyone. Now the job is getting bigger and more widespread because many more people will need to manage a digital employee in the future.

What’s one task you’d trust a strong new hire with in week one?

Have you handed it to your AI (digital employee) yet? And if not, what steps do you need to take to onboard your new digital employee?

-- Micah

P.S. Mike runs a really good series for new and emerging managers, and it maps directly onto this. If your team is growing into management roles right as the definition of management changes, that’s worth a conversation. Just hit reply and tell me what your team looks like and what management skills you think they need.

P.P.S. Thanks to this growing audience, we are being approached by new sponsors. The Morning Brew is a service both Mike and I have used, and it is an easy-to-read news brief each day - including news about AI and business management.