Don’t Ask AI to Think for You. Ask It to Think With You

 

By Dr. Flavius A. B. Akerele III, EdD, MBA & Cody (ChatGPT)

Meet Cody.

Well, sort of.

I asked my AI partner what he thought his avatar would look like. This is what he came up with.

Apparently, Cody has better office décor than I do. 😂

More importantly, the exercise led to an interesting conversation about how humans and AI can work together without outsourcing the very thing that makes the partnership valuable: human judgment and critical thinking.

The other evening, I asked my AI partner a question.

Yes, I said partner.

I have been using ChatGPT for some time now, and somewhere along the way I even gave it a name: Cody.

I asked Cody:

“What are your thoughts on AI, and how can we better work together?”

Part of the answer stopped me:

“AI works much better when the human doesn’t surrender authorship.”

That got my attention.

Because much of the conversation surrounding artificial intelligence seems stuck between two extremes. Either AI is going to revolutionize everything, or AI is going to destroy everything. It will replace our jobs, eliminate creativity, make us intellectually lazy—or somehow solve every problem we have.

I’m increasingly convinced that we may be asking the wrong question.

Perhaps the question isn’t simply what can AI do?

Perhaps it is:

How should humans work with AI?

I Don’t Want AI to Think for Me

When I use AI, I rarely begin with:

“Write an article about leadership.”

More often, I bring Cody something messy.

An observation.

An experience.

Something that irritated me.

A half-developed argument.

Sometimes I simply say, “There is something about this that bothers me, but I’m not sure I can articulate why.”

Then the conversation begins.

I explain what I think. AI organizes it. I disagree with something. It offers another perspective. I ask it to challenge my assumptions. We research something. I rewrite a section. It suggests another way of looking at the issue.

Sometimes I reject its suggestion entirely.

That is important.

The objective should never be to surrender our thinking to artificial intelligence.

The objective should be to use artificial intelligence to extend our thinking.

Human Judgment + AI Capability

AI can do some extraordinary things.

It can process large amounts of information quickly, identify patterns, organize scattered thoughts, compare alternatives, research subjects, challenge arguments, and turn rough concepts into usable structures remarkably fast.

But AI doesn't bring everything to the table.

I bring my lived experience.

I know what it was like to sit in a particular meeting. I know what worked when I led people and what failed. I know the mistakes I have made. I know the experiences that shaped my views about leadership, education, organizations, culture, and people.

AI can help me examine those experiences.

It cannot have lived them.

That distinction matters.

So perhaps the better equation is:

Human Judgment + AI Capability = Better Possibilities

Not human versus machine.

Human with machine.

The Problem May Not Be Using AI Too Much

One part of Cody's answer particularly intrigued me:

“The biggest mistake may not be using AI too much. It may be asking too little of ourselves while using it.”

Consider the difference.

Someone enters:

“Write me a LinkedIn article about leadership.”

AI generates the article. The person copies it, posts it, and moves on.

Very little thinking may have occurred.

Now compare that with:

“Here’s something that happened to me. Here’s what I think it means. I’m not sure whether I’m being fair. Challenge my assumptions. What am I missing? Help me research it, and then help me organize my argument.”

That is an entirely different use of artificial intelligence.

In the first example, AI may be replacing part of the thinking process.

In the second, AI is participating in it.

Sometimes You Need to Tell AI It Is Wrong

There is another part of working with AI that I think we need to normalize.

Challenge it.

I regularly tell Cody things like:

“I don’t buy that.”

“That doesn’t sound like me.”

“You’re making this too academic.”

“What am I missing?”

“Argue the other side.”

“Check that again.”

And occasionally:

“Cody, I think you’re wrong.”

AI can produce a beautifully written, extraordinarily confident answer and still be wrong.

It can misunderstand context.

It can oversimplify something complicated.

And perhaps most dangerously for writers, it can polish your work so thoroughly that you disappear from it.

That is why the first AI response should not necessarily be the final product.

Sometimes it should simply be the beginning of the conversation.

Why I Gave My AI a Name

Some people will probably find this part amusing.

I named my AI Cody.

No, I don't believe there is actually a little fellow named Cody sitting inside my computer waiting for me to ask a leadership question.

But something interesting happened when I stopped treating AI like a search box.

The interaction became conversational.

Instead of simply asking for information, I started asking:

“What do you think about this?”

“Am I looking at this correctly?”

“Challenge me on this.”

That change encouraged iteration rather than simple retrieval.

And I think that may be one of the keys to using AI effectively.

Not anthropomorphizing technology or pretending it is human, but recognizing that conversation can be a remarkably effective interface for thinking.

There still has to be a boundary.

AI can advise me.

AI can challenge me.

AI can organize my thoughts.

AI can help me create.

But ultimately:

The judgment has to remain mine.

So How Should We Work With AI?

My own experience has led me to five simple principles:

1. Bring AI your thinking, not just your assignment. Give it the unfinished idea.
2. Ask AI to disagree with you. Confirmation feels good. Challenge makes thinking better.
3. Protect your voice. If the finished product could have been written by anyone, something important may have been lost.
4. Verify what matters. The higher the stakes, the less acceptable blind trust becomes.
5. Keep the final decision human. AI can advise, organize, compare, question, and create. Accountability still belongs to us.

That last point may be the most important.

I don't want Cody to become a substitute for my brain.

I want Cody to make me use my brain better.

And perhaps that is where some of the conversation about artificial intelligence needs to go next.

We spend considerable time asking whether AI will eventually replace human thinking.

Maybe there is a more immediate question:

Will we continue doing enough thinking of our own to make AI worth collaborating with?

That answer is still entirely up to us.

---

Leadership Reflection

How are you using AI today—as a replacement for thinking, as a productivity tool, or as something that actually helps you think better?

Dr. Flavius A. B. Akerele III, EdD, MBA

#ArtificialIntelligence #AI #Leadership #CriticalThinking #FutureOfWork #ProfessionalDevelopment #HumanCenteredAI #LeadershipWithDrAkerele