Now that millions of hours of online video have been tagged and described by users, new machine-based new-content-creation software, VideAI, has become possible. VideAI is an Artificial-Intelligence-based video editor. A user inputs text and receives a video clip within minutes chosen by the machine to fit that text (the video-clip source is currently YouTube). A machine-learning linguistic algorithm analyzes the user's feedback over time to continually improve the best fit between text and video for future inquiries. VideAI uses YouTube’s metadata and comments to facilitate unsupervised production of new media objects. Combining natural-language processing with video-editing techniques, VideAI allows the user to collaborate with Artificial Intelligence in the creation of new media objects. In this way, the gap between text and image will become smaller and smaller.