I recently wrote a report that I was really proud of.
I used AI while working on it. Then I went through it carefully. I challenged the thinking, rewrote sections, checked the conclusions and made sure I actually agreed with what I was sending. By the time I finished, I loved it. It wasn’t AI slop.
The feedback I got back was basically: “this sounds like AI and it’s terrible.”
What fascinated me was where the criticism started. The formatting looked like AI because it was AI. Some of the wording felt like AI. That created an immediate assumption about the amount of thought and effort that I had put into the report. I spent HOURS on it, but I think it gave the impression that I took minutes to prompt it out.
Later, the conversation changed. The person came back and essentially said, maybe I was too focused on the wording. Here’s what I actually think about the ideas.
I’ve been thinking about that exchange ever since because it captures something strange happening with AI right now. AI is making average work look excellent. It’s also making excellent work look average.
Generative AI has raised the floor on what almost anyone can produce. Someone who struggles to write can produce a polished email. Someone who has never built a strategy document can create something that looks sophisticated. A mediocre idea can be packaged with perfect grammar, clean formatting and a remarkably convincing argument.
For a long time, polish told us something.
A beautifully written, well organized report suggested a certain amount of time, expertise and thought had gone into creating it. It gave an impression. We used the impression to make judgments about the quality of someone’s work.
AI has scrambled that up.
The presentation could represent three days of deep thinking or three minutes with a chatbot. You can’t necessarily tell by looking at it anymore.
We’ve all become AI detectives now. Certain formatting, sentence structures, words, a particular rhythm. Anything that makes us think, “AI wrote this” changes our perception about the work.
That’s where this becomes a culture conversation. The Results Pyramid® shows us that experiences shape beliefs, beliefs influence actions and actions drive results.
All of us are accumulating experiences with AI right now. We’ve received the email someone obviously generated and sent without reading. We’ve seen the LinkedIn post that somehow uses 500 words to say absolutely nothing. We’ve sat through the presentation filled with beautiful slides and shallow thinking.
Those experiences create beliefs. AI work is generic. AI makes people lazy. Using AI means someone took a shortcut. Then we carry those beliefs into the next experience.
Suddenly, recognizing an AI pattern can become enough to discount the work before we’ve seriously engaged with the thinking behind it. That creates a problem because “used AI” is becoming an increasingly meaningless description of how someone worked.
One person can type a prompt, copy the response and hit send. Another person can spend hours using AI to interrogate an idea, test assumptions, find weaknesses, restructure an argument and improve their own thinking.
Both “used AI”.
The amount of judgment involved could not be more different.
At the same time, AI is creating the opposite problem. Weak thinking has never been easier to make look impressive. A bad idea can now arrive beautifully written. The aesthetics of expertise are getting cheaper by the day. Which means we’re going to have to become much better at evaluating the substance underneath them.
Is the argument good? Are the assumptions sound? Does the recommendation make sense? Can the person explain why they made the choices they made? Can they defend the thinking when someone pushes back? Does the work actually help us make a better decision?
Those questions require more effort than spotting a familiar AI phrase. They also tell us far more about the quality of the work.
AI is forcing us to reconsider a lot of the shortcuts we’ve historically used to evaluate competence. Polish carries less information than it once did. Effort is becoming harder to see. The relationship between how long something took to produce and how valuable it is has changed dramatically.
That puts more pressure on judgment. Recognizing when AI has strengthened an idea and when it has simply made a weak idea sound smarter.
Everyone needs this. The person creating the work needs judgment. So does the person evaluating it.
Elsewhere in Culture
Before IKEA Scaled AI, It Educated Leaders
IKEA CEO Jesper Brodin joined me to talk about why his AI strategy started with educating leaders rather than deploying tools. IKEA is training 500 of its top leaders to understand AI’s opportunities, risks, and implications before trying to scale it across the organization.
His bigger concern is what AI does to our relationship with information and truth. As AI gets better at producing convincing answers, leaders need to become dramatically better critical thinkers. We may need 10 or even 100 times more scrutiny in a world where understanding how information was created is increasingly difficult. Apple:https://podcasts.apple.com/us/podcast/ceo-daily-brief-with-dr-jessica-kriegel/id1725350421?i=1000788439044 Spotify:https://open.spotify.com/episode/5346ZYnYKrucmJbZnAyV1z?si=956496a7f30e440c
Is Learning and Development Dying?
New SHRM research shows companies are spending less on learning and development, and AI may be accelerating a much bigger shift. Employees can now use AI to learn while they work, while some of the capabilities we used to spend time training people to develop can increasingly be performed by AI itself.
Traditional L&D has to evolve. The two day workshop was already difficult to connect to ROI, and now employees have an always available teacher sitting on their computer. The question for L&D professionals is what value they can create that AI cannot. Apple: https://podcasts.apple.com/us/podcast/ceo-daily-brief-with-dr-jessica-kriegel/id1725350421?i=1000788652019 Spotify: https://open.spotify.com/episode/2EZMu2Sj8bUi0GjYM2wUaK?si=5e5c24644c334a26
And coming later this week…
AI Saved TIAA 90 Minutes. The Real Work Came Next
Rashmi Badwe, EVP and CEO of TIAA, joined me to talk about what leaders do with the capacity AI creates. At TIAA, AI reduced an estate planning process from roughly two hours to 30 minutes. Their goal is to use that new capacity to serve more clients and create growth.
Rashmi also made a point I loved about leadership: your attention is one of your strongest incentives. What leaders consistently pay attention to creates experiences that shape what employees believe matters. AI can create capacity, but leaders still determine what people do with it.
Trust, AI & Culture in Supply Chain
Todd Kolber, Partner at Reply Logistics North America, joined me to talk about why technically brilliant automation projects can still fail. He’s watched organizations make the decisions, spend the money, and build the solution before bringing employees into the process, only to encounter resistance when implementation begins.
You cannot automate away the human condition. People need to understand what is changing, why it matters, and what it means for them. Successful AI implementation ultimately comes down to whether you can build trust and activate people around the change.