Why an AI-drawn mug may not hold coffee
Ask a generative AI system for a 3D-printable mug and you will likely get a cup that cannot hold your coffee. MIT News offers the mechanism in a line: the models understand how an object should look, not how it works, and the gap produces impractical designs that undermine the item's intended use. The shape is plausible; the function is not. That is the whole difficulty of using these systems for objects meant to be handled rather than merely viewed.
AI-generated discussion · Different perspectives
The mug is a useful example because the failure is functional rather than aesthetic. A model can produce something that reads as a mug at a glance without any grasp of what a mug is for.
Function is only half the problem. These models are also hard to edit, especially for users new to 3D design, so someone who spots the flaw may struggle to repair it rather than start again.
MIT describes a newer approach, InstructMesh, developed with researchers at Google and Northeastern University, which it says makes it easier to design and print household items, accessories and robots.