Multimodal AI processes or generates more than one type of information, such as text, images, audio, video, sensor readings, or actions. A system might answer questions about a chart, create an image from a description, transcribe speech while recognizing speakers, or guide a robot using visual and language input. Models align different modalities in shared representations so information from one form can influence another. This creates flexible interfaces but also introduces new failure modes: visual content may contradict text, timing may be misaligned, and harmful information can cross between modalities. Evaluation must cover each input type, their combinations, accessibility needs, permissions, and the real conditions in which the system operates.
OpenAI has introduced Sora 2, a next-generation video AI model with synchronized audio, improved physics, and tighter control, alongside a companion social app.
Opera today introduced Neon, an AI-powered browser that can run tasks and navigate pages autonomously — aiming to shift web browsing from passive to “agentic.”
OpenAI’s next version of its Sora video generator will default to including copyrighted content unless owners explicitly opt out – drawing criticism from media creators.
AI search startup Perplexity has reportedly raised $200 million at a $20 billion valuation, just two months after a $100 million round – signaling its rapid rise as a major rival to Google.