Nvidia scientists declared ‍Magic3D on Friday, an artificial intelligence model that can produce 3D models from textual explanations. Through inputting a textual stimulus, such as “a ⁢koala dressed‍ in formal attire,” Magic3D produces a 3D mesh model with vibrant texture in approximately 40 minutes. These 3D models‍ could be employed by both video game developers and computer-generated​ imagery directors.