Did he really say that?

The answer is like any question about rapidly evolving technology: it depends on how you use it.

AI does not and cannot replace education.

One of our inspectors under consideration to be Peer Reviewed, who is approaching the 15 years requirement in the inspection trade,  recently shared the following experience after initially consulting with us.

First, full disclosure: he conducted an inspection for a retailer. The retailer wanted to resolve the issue responsibly and simply requested some time to resolve the denial of the claim submitted to the manufacturer.

The homeowners refused, hired a lawyer.

The situation involved a manufacturing defect. However, the manufacturer rejected the claim because the installer had not taken the exact number of required pre-installation moisture measurements.

Notably, there were no visible moisture-related issues—no gapping and no cupping. All moisture content measurements by the inspector averaged 8%. The installers moisture content measurements prior to installation of the OSB subfloor was low for this region, between 8% to 9%. The flooring was red oak strip flooring, This was within the industry accepted range for strip flooring.

That, however, is not the main point of this story.

The cost of the flooring was high enough to move the case from magistrate court to civil court. Both parties agreed to meet in the judge’s chambers in hopes of reaching a resolution.

Our inspector, who has attended several of our Wood Science classes for flooring professionals, called us for advice. We reviewed the opposing inspector’s report and, within five minutes of research, determined it had been generated using AI.

We prepared a list of ten questions based on the report. We advised our inspector to be ready to answer questions concerning his report as well. The retailer’s lawyer was provided with the questions, answers, and supporting sources.

Meeting Day With The Judge

This is important, you must read the entire story to understand the surprise ending.

Both lawyers and inspectors, the retailer’s and homeowners, met with the judge. The retailer’s lawyer questioned the opposing inspector’s expertise. Although the homeowner’s lawyers objected, the judge allowed the questioning, noting that experts should be prepared to be evaluated.

The retailer’s inspector was asked to step outside. (our guy)

The opposing inspector was then questioned—and answered 8 out of 10 questions incorrectly. The lawyer clarified that all questions were based directly on the content of the inspection report he allegedly wrote and submitted as his work as an expert.

The retailer’s inspector was brought back in. When invited to question him, the homeowner’s lawyer had no prepared questions.

The judge then asked our inspector to explain the correct answers. He did so thoroughly, going beyond the answers to explain the underlying science and its relevance to wood flooring.

While it is fair to note he had prior knowledge of the questions, his explanations demonstrated clear subject-matter understanding.

The judge then asked the homeowners inspector who had written the report since he had demonstrated he had no understanding of the subject matter.

The inspector reportedly stated—confidently and somewhat smugly—that he had collected the data according to his inspection school’s classes,  and then used an AI program to write the report.

We understand (though we are not lawyers) that AI-generated reports are often not accepted in court. This case illustrates why.

If an inspector uses AI to assist in report writing, they must still have the knowledge to verify its accuracy and defend its conclusions. They must be able to stand behind their work. An educated inspector writes an outline, then has AI edit the outline before the final version is written. 

AI does not replace education.

The two inspectors were asked to step outside. Less than five minutes later the bailiff asked “our guy” to come back inside but homeowner’s inspector was excluded. The judge ruled that he did not qualify as an expert, and his report was rejected with prejudice. He would not be reconsidered as an expert in the case.

To be fair, that inspector followed the process he had been trained to use—one that  prioritize report production over deep understanding. He had also reportedly been discouraged from pursuing additional education outside his inspection school and school’s association.

This raises a broader issue.

Why discourage inspectors from learning the basic science behind their work—especially when that science is not difficult to understand?

In parts of the flooring industry, there appears to be resistance to well-informed inspectors. An informed inspector is harder to influence, control, with misinformation. An informed inspector reports facts, not industry myths.

And that is exactly what the industry should be striving to eliminate, the reporting of Industry Myths as Factual..

What is important here? This story, after the facts were given to us, was written as a basic outline, Then this story’s OUTLINE was re-written by AI, and minor adjustments made to it. Then the final version rewritten again by us.

I got to admit, it is more concise and attentive to the story, without my “old man” rumblings and ramblings and rants and raves.

And that, is how AI should be used by informed professional flooring inspectors, as a tool to supplement the inspector report writing skills. Not to replace education and understanding of the subject matter. Not to replace writing the report.

Jason Ramsey and John Paul Viveiros are inspectors charging into this brave new world of AI research and report writing, leading the advantcement into how to use AI correctly.