A model is a checkpoint.

The first chair in this experiment now has a CAD model and a set of renders. The form, proportions and overall intent have made it across from the AI-generated reference. The result looks much closer to something that could exist.

But this is still dummy CAD.

I haven’t resolved the structure, joinery, load paths or manufacturing. Those aren’t small details waiting to be added to a finished design. They will help determine what the design can actually be. A clean model and a convincing render don’t make those decisions for me.

That distinction matters especially here because I don’t have direct furniture design experience. I can interpret the image and develop the form in CAD, but knowing how to represent the chair isn’t the same as knowing how it should be built. This exercise brings me up against the edge of what I already know.

Getting the idea into CAD gives me a form I can now question in more detail. The next step is to work backward from it. How would it support someone’s weight? How could the parts come together? Which choices in the image would survive the demands of construction, and which would have to change? The answers may take the chair somewhere the reference never anticipated.

I’m sharing this as a learning exercise, with those questions still open.

This model is a checkpoint between an image I can interpret and an object I can understand well enough to make. Knowing what remains unresolved tells me where the work needs to go next.