The meshes have room for improvement but look promising.
In December OpenAI, the developer of DALL-E and ChatGPT, announced a new AI model powered by Point-E, the company's system for generating 3D point clouds from complex prompts. Some developers have already started experimenting with it, integrating it into different apps. Here is one such experiment: TREE Industries presented some text-to-3D models generated in real time under 20 seconds in Gepetto AI – a technology layer aiming to create 3D worlds by speaking. It can now generate AI for avatars, 3D worlds from text prompts, and more.
The video demonstrates Point-E and Gepetto AI working together to make a dog, a bottle, a lamp, and a motorbike. TREE Industries is planning to use an image prompt at runtime to create the mesh and convert the vertex colors to a material.