July 31, 2026

AI Over 40 Series - Week 30: AI Adoption Barriers: When Information Isn’t Enough

A few months ago, I suggested to my wife that we could use AI to help us communicate better with each other.

It got shut down fast and hard.

I want you to sit with that for a second. Not because the rejection surprised me. It did not, not really. What stayed with me was what it revealed about an assumption I had been carrying through 29 weeks of this series, dozens of client conversations, and nearly every interaction where I had tried to move someone toward AI adoption.

I had the information. I had the evidence. I had 29 weeks of published documentation proving that AI could do exactly what I was proposing. But the proposal itself was the problem.

Not because it was technically wrong. Because suggesting our marriage communication needed a technological intervention said something about us that no amount of capability demonstration could overcome.

The barrier was not information.

The barrier was what the information implied.

The most knowledgeable person in the room

Twenty-nine weeks of building AI literacy have made me more knowledgeable about what is possible than ever before. I can prototype solutions in hours that would have taken weeks. I can research strategic questions that would otherwise require consultants. I can build tools that did not exist, and I have, from the Outlook-Todoist integration in Week 18 to the Career Development Partner in Weeks 27 and 28.

AI has made me genuinely more capable.

Somewhere along the way, I made an assumption that felt so obvious I never really questioned it. If I could just share what I had learned, show people the evidence, demonstrate the possibilities, and document the transformation, they would get excited too. They would start exploring. They would see what I see.

In many ways, I am writing a 52-week series based on that assumption.

And yet the two people with the most front-row access to my transformation, my wife and my daughter, remain largely unmoved.

Not because they lack information. They have had more information about AI’s practical value than almost anyone in my audience. They have watched me build things, solve problems, and transform how I work. They have seen the evidence up close for months.

It did not matter, because information was never the barrier.

Two barriers information can’t touch

My wife has used AI. She has found it helpful at times for personal tasks. She is not uninformed, and she is not resisting out of ignorance. She has firsthand positive experience with AI tools, and yet her fundamental position has not changed. She still sees reliance on AI as a form of cheating, as taking a shortcut past the effort that makes the work legitimately yours.

That is not an information gap. That is a values position about authenticity and earned effort. More information does not weaken that concern. In some ways, it reinforces it. Every time I demonstrate something impressive that AI helped me create, it confirms the very thing she is worried about: look how much of the work the machine is doing.

The evidence I think should persuade her is the evidence she finds most troubling.

My daughter’s barrier is different. She is a teacher, and her professional identity is built around facilitating the productive struggle that creates real learning. When she watches students reach for AI instead of working through problems themselves, she is not seeing a useful tool. She is seeing a threat to the process she believes creates genuine understanding.

Her concern about what happens to learning when AI does the thinking is not resistance to technology. It is a legitimate pedagogical question educators are wrestling with right now, and she is much closer to the front lines of that question than I am.

When I showed her impressive AI capabilities, I thought I was opening a door. In reality, I was threatening the foundation of what she has built her career around. Every demo that made AI look more capable made her more concerned, not less.

Two people who love me. Two people with unlimited access to my evidence. Two completely different barriers.

Neither one responsive to more information.

The trap

Here is what I have been forced to confront: the assumption that better information produces better outcomes with people may be one of the most dangerous things AI literacy teaches you.

Think about what AI does extraordinarily well. It gives you information, analysis, research, frameworks, strategies, arguments, and evidence. It makes you faster at finding data, more thorough in your analysis, and more articulate in your presentation.

Every one of those capabilities reinforces the instinct that the path to change runs through better information.

For a certain category of problems, that is exactly right. When someone is stuck because they do not know something, and when the barrier is genuinely informational, AI is transformational. The contract routing solution in Week 10 needed information about what was possible with Microsoft Forms and Power Automate. The AI Automation Decision Architecture in Week 26 needed a systematic comparison of tools and capabilities. In those cases, information was the barrier, and AI helped remove it.

But when the barrier is values, identity, trust, fear, or genuine disagreement, more information does not just fail to help. It can actively make things worse.

When you show up with better data and more compelling arguments in response to a concern that is not really about data, it can feel like you are not listening. It can feel like you have diagnosed the other person’s position as an information deficiency rather than a legitimate perspective.

I did this with my wife for months. Not aggressively, because I am not that oblivious, but persistently. A steady stream of “look at this cool thing AI can do” moments. Casual demonstrations. Gentle suggestions. All of it calibrated to build the case through accumulated evidence.

If I am honest, it was a persuasion campaign. A sophisticated one, maybe, but still a persuasion campaign. And it was built on the assumption that the right demonstration, delivered at the right moment, would eventually shift her position.

It did not, because her position was never about what AI can do.

It was about what relying on AI means.

What actually worked

The thing that changed the dynamic in my household was not a better demonstration, a more targeted use case, or the perfect problem that AI could solve for someone else. It was that I stopped trying so hard to convince them. I stopped performing the role of AI evangelist in my own home. I stopped treating family conversations as potential conversion opportunities. I stopped curating moments designed to shift their thinking. Instead, I started paying attention to what they actually cared about.

My daughter has a specific student she has been working hard to find solutions for, a genuinely difficult teaching challenge. I offered to help her think through approaches, and yes, AI was part of how I could help. But the conversation was not about AI. It was about her student.

It was not, “Let me show you what AI can do.”

It was, “Tell me about this kid, and let’s figure this out together.”

That opened doors months of AI advocacy never did. Not because the information was better targeted, but because the pressure was gone. I was not trying to convince her of anything. I was trying to help her with something she cared about. AI happened to be useful in that effort, but it was not the point of the conversation. The most effective thing I did was stop trying to persuade.

That is uncomfortable for me to admit. I build frameworks for a living. I write a weekly series documenting evidence-based transformation. I created a diagnostic process modeled on medical differential diagnosis. My professional instinct is to analyze, strategize, and optimize.

And the breakthrough with the people closest to me came from not doing any of that.

AI cannot really teach you that. I could spend hours with Claude strategizing about how to persuade my wife or convince my daughter, and every strategy would likely become another version of the same mistake: treating a human relationship as a problem to be optimized.

The organizational mirror

If this only applied to family relationships, it would be an interesting personal essay and nothing more. But I have watched the same pattern play out professionally.

In Week 14, I described the moment our organization shifted from “some people use AI” to “everyone uses it.” I framed it as moving from encouragement to expectation. But looking back, the breakthrough was not simply the expectation. It was a developer who said, “I do not have time to learn AI,” tried it, and called the next day, saying what he learned in a single day transformed everything.

I did not persuade that developer with information. I created conditions where he could have his own experience.

There was visibility into who was using what. There were direct conversations about specific objections. There was access to current models instead of outdated free tiers. And yes, there was ultimately clear expectation setting.

But then I had to step back and let the experience do the work.

The information I had been sharing for months did not move him. His own experience in a single day did.

In Week 15, I identified the seven barriers to process improvement, the reasons organizations do not fix broken processes even when the technology exists to do so. Every one of those barriers is fundamentally human, not informational. We no longer see the problems. Nobody owns them end-to-end. Fixing them feels too expensive. The current process is “good enough.”

Those are not just knowledge gaps. They are psychological, political, and organizational barriers that more information alone cannot solve.

I knew this. I wrote about it. And I still walked into my own kitchen, assuming that the right information would change my wife’s mind about AI.

The question I have to ask about this series

This is the part that is genuinely uncomfortable. What is this series if not a 52-week information strategy?

I set out to document my AI transformation, share evidence, build frameworks, and demonstrate possibilities, all with the underlying assumption that if leaders had better information about AI, they would transform too. Week after week, I have been writing the most comprehensive personal AI case study I can and presenting it as the path forward.

My own family, the people with the most direct access to that information, proves the assumption is incomplete at best.

Does that mean the series is wrong? No. Information matters. For people whose barrier genuinely is, “I do not know where to start,” or, “I do not understand what is possible,” this series provides real value. For the leader treating AI like Google, waiting for the obvious use case, or struggling to distinguish agent hype from reality, information may be exactly what they need.

But for the leader whose real barrier is, “This threatens how I see my role,” or, “I do not trust something I cannot fully understand,” or, “I have been successful for 25 years without this, and now you are telling me that is not enough,” this series by itself will not move them.

No amount of well-documented evidence will. Because their barrier is not information. It is something deeper. And treating it as an information problem, even a sophisticated, empathetic, 52-week information problem, misdiagnoses the challenge.

That may be the most humbling realization of this entire series so far: I am 30 weeks into writing about AI transformation, and the most important thing I have learned about transformation has nothing to do with AI.

What AI actually gives you for the hard human work

So if AI does not give you the power to persuade people through superior information, what does it give you?

It gives you diagnostic capability. Not in the sense of, “Here is how to convince this person,” but in the sense of, “Here is what the actual barrier might be.” When I finally stopped trying to sell AI to my wife and started trying to understand why she resisted it, the conversation changed. AI could have helped me think through the difference between a values barrier, an information barrier, and an identity barrier, and that diagnostic clarity would have been genuinely useful.

It also gives you preparation capacity. When my daughter described the student she was struggling with, I could research approaches, synthesize strategies, and prepare thoughtful suggestions faster and more thoroughly than I could have without AI. The AI did not do the relational work of earning her trust or showing genuine interest. But it made me better prepared for the conversation once the door was open.

AI can also give you self-awareness tools. One of the hardest things about the knowledge trap is recognizing when you are in it. AI can help you examine your own assumptions, identify where you are defaulting to information strategies, and consider what the actual barrier might be.

The irony is that I had to learn this lesson the hard way before I could articulate it clearly enough to ask AI to help me watch for the pattern.

But here is what AI does not give you: the patience to stop strategizing and simply be present with someone. The humility to accept that their resistance might be legitimate rather than a problem to solve. The discipline to stop solving and start listening. The willingness to let people arrive at their own conclusions on their own timeline.

That is the hard human work.

AI gives you better tools and ideas to bring to that work. It does not do the work for you. And the most important part of the work, the part where you stop trying to persuade and start trying to be genuinely helpful, may be the part AI is least equipped to teach.

The framework underneath

I have started thinking about barriers to change in three categories that go beyond the simple information-versus-emotion split I started with.

The first category is value barriers. This is when the person understands what you are proposing but opposes it because it conflicts with something they believe matters more. My wife values authenticity and earned effort. AI challenges that value. More evidence of AI capability reinforces her concern rather than addressing it.

The second category is identity barriers. This is when the person understands what you are proposing but fears it because it threatens how they see themselves professionally or personally. My daughter’s identity as a teacher is built on facilitating the struggle that produces learning. AI does not just challenge her workflow. It challenges her understanding of what her job fundamentally is.

The third category is interest barriers. This is when the person understands what you are proposing and resists because it genuinely threatens something they value, such as status, autonomy, comfort, or competitive advantage. This is not ignorance, and it is not merely emotion. It may be rational self-interest. No amount of persuasion fixes that by itself. It requires negotiation, compromise, or structural change.

The diagnostic question is simple: when someone is not moving, ask yourself what kind of barrier you are actually facing.

Is the barrier that they do not know? Does that conflict with what they believe? Does it threaten who they are? Or that it threatens what they have?

Each barrier requires a fundamentally different approach, and only the first one responds reliably to better information. The mistake AI-literate leaders are most likely to make is treating every barrier as the first type, because that is the barrier AI helps with most naturally.

Your Week 30 challenge: Find your knowledge trap

This week, identify someone you have been trying to move with information.

It might be a colleague, a team member, a client, or a family member. Think of someone you have shared evidence with, demonstrated possibilities to, or made the case for, and yet they still have not moved.

Then stop and ask: what if information is not their barrier?

Try to diagnose the actual barrier. Is it values? Identity? Interest? Trust? Fear? Something else entirely? What would you learn if you stopped presenting evidence and started asking questions about what concerns them most?

Also notice where you are persuading, or at least trying to. Catch yourself in the act of being the knowledgeable one. Are you sharing information because it is genuinely what the other person needs, or because being knowledgeable feels comfortable and AI makes it easy?

Then try the thing that worked for me. Pick one relationship where you have been advocating for something and stop advocating for a while. Find out what that person is actually struggling with right now. Help with that.

Let the tool you believe in prove itself in service of their problem, not yours.

Finally, ask yourself the uncomfortable question about your own strategies. Where have you built an elaborate information-based approach to changing someone’s mind? What might happen if you stopped optimizing and started listening?

The bottom line

AI has made me more knowledgeable than I have ever been. It has not automatically made me more persuasive. For months, I confused the two.

Knowledge is power, but power is not persuasion. The ability to research, analyze, synthesize, and articulate faster than ever before is genuinely transformational when the barrier you are facing is informational. When it is not, all that capability can actually work against you. It makes you more likely to reach for information-based solutions. More confident in your evidence. More persistent in your case-making. And less likely to notice that the person across from you stopped listening three demonstrations ago.

AI transforms what is possible. It does not transform how quickly humans accept new possibilities. That is not a limitation of AI. That is part of being human. The hard human work remains hard: building trust, respecting values you do not share, earning the right to be heard, and letting people arrive at their own conclusions on their own timeline. AI gives you better tools to bring to that work. It does not do the work for you. And sometimes the most important insight is knowing when to put the tools down entirely.

Thirty weeks in, the most powerful thing I have learned about AI has nothing to do with prompting, infrastructure, or frameworks. The people closest to me taught me more about the limits of knowledge than any research paper ever could.

This post is part of my “AI Over 40” series. It first appeared on LinkedIn: AI for the Over 40 [Week 30]: Knowledge Is Power. Power Isn’t Persuasion.

Read more AI and Copilot blogs.

Trending Posts