Amazon develops artificial intelligence model for video analysis

Amazon is developing an artificial intelligence video model that will allow users to find specific scenes in a video with text commands, a data source reports.

Last week, Amazon announced a new $4 billion investment in Entropy, maker of the Claude chatbot, bringing the retail giant’s total investment in the AI ​​company to $8 billion, but according to a new report published by The Information, Amazon independently An artificial intelligence model is being developed for video and photo analysis.

With Amazon’s artificial intelligence, you can find specific scenes in the video

This artificial intelligence model that Amazon is currently developing is called Olympus and it is multifaceted. Multimodal AI can process different types of data such as image, text and video.

Amazon apparently plans to use Olympus for video analytics. How exactly this artificial intelligence works is not yet clear, but The Information says that users can find certain scenes or details of the video with the help of Olympus just by using a simple text command.

Anthropic, which has so far attracted more than $8 billion in funding from Amazon, unveiled the Claude 3 family of multimodal AI models earlier this year. Now it remains to be seen whether Amazon will use the large multifaceted language models already developed by Entropy, or if it really wants to develop its own proprietary models.

According to published reports, although Amazon is a shareholder of Entropic, it plans to reduce reliance on the company for AI models. The time of the unveiling of Olympus artificial intelligence is not yet known, but Amazon will hold the AWS re: Invent 2024 conference in the next few days, and we may see the unveiling of Olympus video artificial intelligence in this conference.

The collaboration between Amazon and Anthropic may even affect the voice assistant Alexa. This voice assistant is going to receive new capabilities based on artificial intelligence in 2025.