Amazon is developing an Artificial Intelligence (AI) model that will be able to analyze videos and find specific scenes that users have requested via text and which is called Olympus AI according to The Information.
The company is considering launching this technology soon, within the framework of its annual conference AWS re:Invent which will begin next Monday, December 2 and will last until Friday of next week.
Currently, Amazon Web Services (AWS) and Anthropic maintain a strategic collaboration to promote generative AI, working together to develop technologies like Trianium and Inferentia. Additionally, Anthropic provides AWS customers with access to different generations of its foundational models through Amazon Bedrock.
Although a few days ago it was announced that they would strengthen their alliance to continue “working closely to continue advancing Trinium’s ‘hardware’ and ‘software’ capabilities”, as they formalized in a statement, Amazon plans to reduce its dependence from your collaborator for multimodal AI.
Sources related to this technology company have confirmed to The Information that Amazon is working on its own artificial intelligence model, which can process images, videos and text.
This new language model will be called Olympus AI and will allow users to search for video files for specific scenes, as reported by Business Insider, which has had access to this information and points out that
Thanks to this functionality, users will be able to ask the tool to analyze a video of a basketball game and find the specific moment in which the ball is scored, according to The Verge.