Artificial intelligence is everywhere right now—including product photography.
For makers, the promise is appealing: generate beautiful, styled product images instantly, without the time or cost of a photo shoot. But how well does it actually work for handmade, functional pieces?
I decided to find out.
Using one of my own pieces—a handmade ceramic butter keeper in a Starry Night glaze—I tested whether AI could create accurate, usable product images.
Here’s what happened.
What Is a Butter Keeper? (And Why It’s Hard to Fake)
A French butter keeper isn’t just decorative—it’s designed to keep butter fresh and spreadable without refrigeration.
It works like this:
- Butter is packed into the inverted bell-shaped lid
- The base holds a small amount of water to create an airtight seal
- The bell fits into the base, protecting the butter from air
This design depends on precise proportions and correct orientation.
That makes it a perfect test for AI-generated imagery.
The Goal: A Simple Lifestyle Product Photo
I asked AI to create a realistic image of the butter keeper being used at a table—someone spreading butter on toast.
To help, I provided a real reference image of my handmade piece.
Simple enough, right?
Not exactly.
Where AI Product Images Failed
Despite detailed instructions and multiple revisions, the results consistently missed critical details.
1. Butter in the Wrong Location
The most common mistake: butter placed in the base instead of the inverted bell.
This might look acceptable at a glance—but it’s completely incorrect.
2. Incorrect Use of the Base
In some images, the base (which should sit on the table holding water) was being lifted or tilted, as if it contained the butter.
In reality, that would make the piece unusable.
3. Extra or Nonexistent Parts
Some versions introduced lids, knobs, or features that don’t exist in the actual design.
These additions made the piece look generic instead of handcrafted.
4. Broken Proportions
The bell and base often didn’t fit together properly, which defeats the entire purpose of the butter keeper.
For functional pottery, proportions aren’t just aesthetic—they’re essential.
Why AI Struggles with Functional Pottery
The issue wasn’t a lack of detail in the prompt.
The problem is how AI generates images.
Instead of understanding how an object works, AI relies on patterns from training data. In this case, it defaulted to familiar visuals like butter dishes and bowls.
It optimized for what looks right, not what is right.
For decorative objects, that can be enough.
For functional handmade pottery, it isn’t.
Why Accurate Product Photography Matters
If you sell handmade pottery online, your photos do more than showcase appearance—they explain function.
Customers rely on images to understand:
- How a piece is used
- What makes it unique
- Whether it fits their needs
Inaccurate images can create confusion, reduce trust, and ultimately hurt sales.
Where AI Images Can Still Help
AI isn’t useless—it just needs to be used strategically.
It can work well for:
- Lifestyle backgrounds and scene ideas
- Branding and marketing visuals
- Concept exploration for new designs
But for core product images, especially functional items, real photography is still the best option.
Final Verdict: AI vs. Product Photography
After testing AI with a real handmade product, the conclusion is clear:
AI can imitate the look of pottery—but it struggles to represent how it actually works.
And for functional pieces, that difference matters.
The Takeaway for Makers
If you create handmade work, your advantage is authenticity.
Real materials, real craftsmanship, and real functionality are what set your work apart—and your product photography should reflect that.
AI can support your brand, but it shouldn’t replace the real thing.
Looking for handmade ceramic butter keepers?
Explore my collection at Joe Dallas Pottery and see how each piece is designed to be both functional and beautiful.