Amazon is set to unveil its new generative AI model, Olympus, which can process images, videos, and text, during the upcoming AWS re:Invent conference on December 4, 2024.

Olympus is expected to be integrated into AWS, potentially utilizing AWS Bedrock, which already hosts several Amazon-developed AI models.

In 2024, Amazon has committed $75 billion to expand its cloud and AI infrastructure, a significant increase from the previous year's $48.4 billion.

The partnership with Anthropic will enhance Amazon's digital assistant, Alexa, which is expected to receive a major AI upgrade in 2025.

Olympus is designed to assist in various applications, such as analyzing geological data for energy companies and improving search capabilities in video content.

While specific functionalities of Olympus remain undisclosed, its capabilities are expected to compete with existing models in the generative AI market.

This new development aims to reduce Amazon's reliance on Anthropic's Claude chatbot, a key feature of Amazon Web Services (AWS).

Amazon's AI initiatives also include the development of two chip lineups, AWS Trainium and AWS Inferentia, specifically optimized for AI workloads.

The new AI chip, Trainium 2, developed by Annapurna Labs, is anticipated to be operational in Amazon's data centers by the end of 2024, providing a cost-effective alternative to Nvidia's GPUs.

Despite these advancements, Amazon faces stiff competition from companies like Nvidia and Alibaba, which are also making strides in the generative AI space.

Historically, Amazon's focus has been on generative AI models primarily in the text-to-text domain, but Olympus marks a significant expansion into multimodal capabilities.

The information regarding Olympus was first reported by The Information on November 27, 2024, highlighting the anticipation surrounding its launch.