AI Slop or a Tool for Better Work?

Why AI Generated Images Are Not “Slop” When Paired With Real Expertise

Here’s the deal. AI generated visuals get dismissed constantly as low effort filler, and I genuinely understand why that reputation exists in a lot of places online. Used the way I actually use it, to illustrate real construction and Lean concepts I have genuinely lived through, it is something completely different, and I want to explain exactly why.

Why Confidentiality Used to Limit What I Could Actually Share

Some of the best ideas I encounter come from projects covered under a genuine confidentiality agreement, which means I cannot share the real photos or specific details directly. Before AI became a genuinely useful tool for this, that meant a lot of real, valuable lessons simply stayed locked away, unable to help anyone beyond the specific project where they happened.

Now I can take an actual photo, extract the underlying concept it illustrates, and have AI recreate that concept as a new, original image, one that teaches the exact same principle without ever violating the confidentiality agreement or revealing anything genuinely protected. That is not cutting corners. It is finding a genuinely ethical way to share a lesson that would otherwise never reach anyone at all.

This matters more than it might first appear. Some of the most advanced, genuinely impressive implementations of Lean and Takt thinking I have encountered exist specifically inside organizations that cannot let any of it be photographed or discussed publicly. Without a way to responsibly recreate those lessons, that knowledge would simply stay locked behind a confidentiality agreement forever, benefiting nobody outside that one project.

Why Recreating International Examples Helps People Actually Learn

A related problem shows up whenever I share real examples from Japan. People often dismiss them immediately, assuming the lesson only applies there and has nothing to do with their own project here. Recreating the same underlying concept as an original illustration, rather than relying on the original photo alone, often helps people actually engage with the idea honestly, rather than mentally filing it away as something foreign and irrelevant before they have genuinely considered it.

Stripping away the specific location or context, while keeping the actual principle fully intact, tends to force a more honest evaluation of the idea itself. People end up judging the concept on its own merits rather than dismissing it based on an assumption about where it came from.

I have also used this to illustrate real things I did earlier in my career, moments I never happened to photograph at the time, but remember clearly enough to describe accurately. Turning an accurate memory into a genuine visual teaching tool has given me more real content to actually teach with than I have ever had available before.

Think about how much genuinely valuable experience exists in the industry purely as memory, never documented visually because nobody happened to have a camera out at the right moment. That knowledge used to simply disappear over time. Now it can actually become a real teaching resource, as long as the person describing it genuinely knows what they experienced and checks the result carefully against that real memory.

Here are the specific situations where this approach has genuinely proven valuable.

• Recreating a confidential project detail as an original image that teaches the same lesson without revealing anything protected.
• Turning an international example into a more universally relatable illustration, so people engage with the actual concept rather than dismissing it as foreign.
• Visualizing a real past experience that was never photographed at the time, based on an accurate firsthand description of what actually happened.

Why the Difference Comes Down to Genuine Quality Control

None of this works without real expertise sitting behind it. Every image I generate this way gets checked personally against my own firsthand knowledge. I have actually done these things, actually seen them happen, and know directly that what is being illustrated is genuinely possible and accurate. That review process is exactly what separates a genuinely useful teaching tool from the kind of low effort, unreviewed content that earns AI generated work its poor reputation in the first place.

That distinction is genuinely the whole point. Anyone without real domain knowledge can generate an image that looks technically polished but contains details that are subtly or obviously wrong. Someone who has actually spent years in the field can catch those errors immediately, which is exactly why expertise, not the tool itself, is what determines whether the final result genuinely teaches something true.

Watch for these signs AI generated content is actually being used responsibly, with real expertise providing genuine quality control.

• The person generating the image has genuine, firsthand experience with the concept being illustrated.
• Every image gets reviewed carefully before being shared, rather than published exactly as first generated.
• The underlying concept traces back to something real and verifiable, rather than an invented or speculative scenario.

Why This Has Genuinely Accelerated My Own Work

Because of this approach, I have been able to write books at a pace I never thought was possible, research more thoroughly, and pull together material from across a much wider range of real sources than I could have managed working entirely alone. The volume of genuinely useful teaching material I can now produce has grown significantly, and the quality has not suffered, because real expertise still sits behind every single piece before it ever gets shared.

I genuinely believe the volume and the quality are connected rather than in tension with each other. Producing more material faster only works well when the underlying review process stays just as rigorous as it would be for a single, slowly produced piece of work. Skip that review, and speed simply amplifies whatever mistakes slip through, which is exactly the trap worth avoiding.

Why I Think This Technology Deserves a Fair Chance

New tools often face real skepticism before people fully understand how to use them well, and that skepticism tends to fade once the genuine value becomes clear through actual results rather than theoretical debate. I would rather keep using AI thoughtfully, with real expertise and genuine review behind every piece of content it helps me create, than dismiss the whole approach because of how poorly it gets used elsewhere.

That skepticism is not unreasonable either. Plenty of AI generated content genuinely deserves the criticism it gets, precisely because it skips the review process entirely. The goal is not defending every use of the technology indiscriminately. It is making the case that thoughtful, expertise driven use is a genuinely different thing from the low effort version people are rightly frustrated by.

If your project needs superintendent coaching, project support, or leadership development, Elevate Construction can help your teams think through how to use AI responsibly in your own training and documentation, pairing it with real, firsthand expertise rather than treating it as a shortcut around genuine knowledge.

So here is the challenge. If you are using AI to create training content, checklists, or documentation of your own, ask honestly whether genuine expertise is actually reviewing what comes out the other end. That single habit is what determines whether the result is a genuinely useful teaching tool or exactly the kind of unreviewed content people are right to be skeptical of.

Jason Schroeder said it plainly: “I QC every image. I’ve done it, I’ve seen it, I know it’s possible.” That standard, not the tool itself, is what actually determines whether AI generated content is worth sharing.

On we go.

FAQ

Why does using AI to recreate a confidential project detail avoid violating a confidentiality agreement?

Because the recreation illustrates the underlying concept or lesson rather than reproducing the actual protected photo, document, or specific project detail itself. That distinction lets a genuinely valuable lesson reach a wider audience without exposing anything a client or partner genuinely expected to remain confidential.

Why might recreating an international example help people engage with a lesson more than the original photo would?

Because people sometimes mentally dismiss an example as culturally specific or irrelevant to their own situation before genuinely considering whether the underlying principle actually applies. A recreated, more universally relatable illustration can help someone engage with the actual concept honestly, rather than filing it away as something foreign that does not apply to their own work.

What specifically separates responsible use of AI generated content from the kind of “AI slop” people criticize?

Genuine firsthand expertise reviewing every piece of content before it gets shared, rather than publishing whatever AI produces without real scrutiny. Content grounded in something the creator has actually experienced and verified, checked carefully rather than accepted automatically, is fundamentally different from content generated and shared with no genuine review behind it.

Why does the article compare skepticism toward AI to historical skepticism toward earlier technologies?

To illustrate that genuine unfamiliarity with a new tool often produces real skepticism that fades once people see how it can actually be used well, through real examples rather than abstract argument. The comparison is not meant to dismiss legitimate concerns about the technology, but to suggest that thoughtful, expertise driven use tends to demonstrate real value more effectively than theoretical debate about whether the tool itself is good or bad.