What AI Is Actually Teaching Me About My Business

I came across one of those random Facebook discussions asking business owners a few simple questions: Which AI platforms are you using? What did you actually use AI to accomplish in the last seven days? What did you try that didn't work?

That last question got me.

There is plenty of conversation about what AI can do. Depending on who you listen to, AI is either going to save every business in America or destroy civilization before lunch.

I'm much more interested in what it is actually doing inside my business.

Not theoretically. Not what somebody on YouTube says I should be doing. What did I actually use it for?

The Tools

I use ChatGPT heavily. Claude has been involved particularly in some of the technical architecture surrounding the Photography Fit 2 Print provenance system.

Adobe's AI-assisted tools are already woven into my photography workflow. I also use Evoto, Luminar Neo and ON1 tools for different parts of image processing. Magica has been another part of the broader environment I've been experimenting with.

But simply listing platforms doesn't tell you very much.

Owning a hammer doesn't mean you built the house.

The more interesting question is what all these tools are helping me accomplish.

What Did I Actually Do With AI?

During the past week alone, I've used AI to work on Photography Fit 2 Print from several different directions.

I've refined how I talk about the business. We've worked on messaging around printing, restoration, archival storage, NAS systems, provenance and something I believe strongly: photographs deserve a life beyond a screen.

I've developed actual marketing material. Facebook posts. Threads posts. Website articles. Social media teasers. Ideas designed to bring people back to the Photography Fit 2 Print website and, hopefully, help them understand why printing, preservation and restoration matter.

But I'm not just using AI to advertise the business.

I'm using it to help build the business.

I've worked through equipment inventory, storage architecture, file migration, NAS reliability, digital-asset protection and operational documentation.

That distinction matters to me. It is easy to reduce AI to "write me a caption." There is nothing wrong with that. I do it too.

But things became much more interesting when I stopped looking at AI simply as a writing machine and began using it as something closer to a collaborator, analyst and occasionally an intellectual sparring partner.

Then We Started Building Provenance

Photography has entered an interesting period. A photograph can now be captured, computationally processed, retouched, altered, generated or some combination of all five.

That creates a question I believe photographers and clients are increasingly going to care about: what exactly am I looking at?

Photography Fit 2 Print has been developing a provenance system around that question. We have been working through an append-only ledger, XMP-first evidence gathering, cryptographic hashing, asset and master relationships, print records and rules surrounding generative content.

Machines can gather evidence. Machines can propose classifications. But I have established an important boundary:

A machine should not silently make a material claim about the authorship or provenance of my artwork.

There is still a human being responsible for the work. Me.

That principle may eventually prove more important than the technology underneath it.

AI Has Also Been Helping Me Write

I'm also working on a book. AI has helped me with continuity, structure, editing and occasionally figuring out what I am actually trying to say.

But I don't want AI writing a book and attaching my name to it. There is a difference between helping me find the sentence and inventing the experience behind the sentence. Those experiences belong to me.

Sometimes the most useful interaction I've had with AI ends with me saying: "No. That isn't what I'm trying to say."

That may sound like failure. It isn't. Sometimes having something reflected back at you makes your own position clearer.

I Use It in My Other Professional Life Too

Photography is not my only professional discipline. I come from project controls and scheduling, so I've also used AI to examine employment opportunities, review my resume, evaluate how my experience aligns with positions, prepare for interviews and organize CAPM study material.

That background has probably influenced how I'm approaching AI more than I initially realized. Project controls teaches you not to become impressed simply because something is moving. You want to know: What is the scope? What is the baseline? Who owns the decision? What changed? What is the impact? Where is the risk?

AI deserves exactly the same questions.

Now Let's Talk About What Didn't Work

There was a Photography Fit 2 Print campaign poster. It was terrible. Not mediocre. Terrible.

I supplied finished photographs that I considered completed pieces of artwork. During the design process, those photographs effectively became raw material to be manipulated.

I hated the result. We stopped. But something useful came out of it. I established a rule:

My photographs are finished artwork unless I explicitly authorize their alteration.

That failure produced an operating standard. And that might be one of the most important things AI has done for my business.

Failure Can Reveal the Rule

There have been smaller failures too. AI can over-polish my writing. Suddenly something I said sounds technically perfect but doesn't sound anything like me. My humor disappears. My cadence disappears. My lived experience gets replaced by something that sounds like it came from a conference-room presentation. No thank you.

AI can also move toward execution too quickly. The provenance work has reinforced something I already understood from project controls:

Establish the rules before you automate the process.

Authority first. Evidence second. Controls in place. Then automation.

AI also has a tendency to give you fifteen good ideas simultaneously. That sounds wonderful until you realize you haven't fully understood idea number two and you're already saying yes to number eleven.

Again, project controls. Scope it. Sequence it. Understand the dependencies. Then authorize the work.

Maybe We're Asking the Wrong Question

Most conversations about artificial intelligence begin with: what can AI do? I'm beginning to think a more useful question for a business owner might be:

Where did AI fail, and what did that failure teach me about my business?

The failed poster taught me something about artistic authorship. The provenance project taught me something about human authority. The NAS work taught me how easily technology can create a false sense of security. The writing taught me the difference between polish and voice.

Those are not AI tricks. Those are governance lessons.

An AI Policy We Never Intended to Write

One unintended consequence of all this is that Photography Fit 2 Print appears to be developing an AI policy without ever sitting down and declaring, "Today we're going to write an AI policy."

Instead, the policy is emerging from the work. Authorship matters. Provenance matters. Finished artwork should remain finished artwork. Generative content should be disclosed appropriately. Machines can gather evidence and make recommendations, but material decisions require human accountability. Technology should support the creative process rather than quietly taking ownership of it. Archival integrity matters. And just because something can be automated doesn't mean it should be automated.

I didn't begin experimenting with AI expecting to arrive here. I was trying to become more efficient. Instead, the technology has forced me to become more specific about what Photography Fit 2 Print actually believes.

And perhaps that is one of the most valuable things a new technology can do. It forces you to decide where you stand.

We are Fit 2 Capture. We are Fit 2 Curate. We are Fit 2 Print.