I spent yesterday morning staring at two monitors in my home office, dissecting the fine print of some insurance requirements for a job.
My client, an IT managed services firm, was heading into a project at a property in Texas. They sent me two documents—the insurance requirements, and a vendor endorsement form from the property management group. To a casual observer, or perhaps a high-speed AI tool, these documents look like standard “check-the-box” paperwork.
But I realized that if I treated them with the same sanguine detachment many tech enthusiasts have for automation, I would leave my client’s neck on the line.
The $39 Correction
The requirements were dense. They demanded Primary and Non-Contributory coverage, and a Waiver of Subrogation. These are legal protections that prevent insurance companies from bickering over who pays a bill, or suing a third party after a claim is settled.
As I read the documents aloud with an underwriter from The Hartford, a “crooked” detail surfaced: my client’s Workers’ Comp policy was only set for their home state of Arizona.
I felt a sudden stiffness in my chest. To an insurance partner, a miscalculation feels like a personal failure. It turned out the premium increase to add Texas was only thirty-nine dollars. I texted my client, explained the gap, and he shot back, “Sounds great!”
A machine can calculate a premium, but it cannot feel the weight of stewardship. It doesn’t feel what a human agent feels when a client is one accident away from a denied claim, because of sloppy paperwork.
The “Blanket” Trap
Later that afternoon, while driving my son to work in Gilbert, I was back on the phone with The Hartford. They’d sent the “corrected” certificate… but it was still wrong.
The documents specifically required two separate entities to be listed. They’d failed to add the second one.
Here is where the human soul beats the algorithm every time. Most vendors rely on “blanket” endorsements that trigger coverage “where required by written contract.” However, because my client didn’t hold a direct contract with either of these two parties, a blanket endorsement would leave wiggle room for a trial attorney.
Bottom line: the property owners and financiers would have remained unprotected, and if something went sideways, they’d sue my client to make up the difference.
The requirements were clear: they needed these entities specifically listed on a Schedule. With the car quiet and my son focused on the road, I stayed on the line until the underwriter got the names exactly right.
Sharpening the Axe
The next day, I debated with a group of entrepreneurs about the power of AI. They spoke with passion about how much time automation can save. I spoke with equal fervor and quoted Abraham Lincoln: “If you have six hours to cut down a tree, make sure you spend the first four sharpening the axe.”
When it comes to risk, entrepreneurs are in the “axe-sharpening” business. We live in an era where people want to make everything simple as the touch of a button. But as Ecclesiastes says, “What is crooked cannot be made straight, and what is lacking cannot be counted.”
AI lacks a soul. It lacks the “tacit knowledge” to look at a Texas requirement and cross-reference it with an Arizona policy. It lacks the intuition to look for a missing name on a schedule when the system says “Complete.” Often, it takes much longer to anticipate all these scenarios, and program AI to respond to them, than to simply do the work yourself.
I emailed the final certificate that afternoon with a deep sense of peace. My client was protected. The entities were scheduled. The axe was sharp.
Technology is a fine tool, but when the financial life of a business is on the line, you don’t need a faster processor. You need a human being who cares enough to stay on the line until the job is done right. If you want someone like that in your corner, maybe we should talk.
Leave a Reply