Multimodality
Multimodality refers to the capacity of models to process text, image, audio, and video together. Search itself is becoming multimodal: photograph, speak, show — rather than type.
In practice
The editorial consequence: informative images — diagrams, tables, real labeled photographs — become first-class content, understood and citable by models. Optimization returns to basics (descriptive filenames, alt text, context) with a new question attached: do your visuals answer a question, or merely decorate?