Home ยป Amazon Engages in Pioneering AI Apprenticeship to Outsmart OpenAI

Amazon Engages in Pioneering AI Apprenticeship to Outsmart OpenAI

Amazon, the e-commerce giant, is reportedly investing in the development of artificial intelligence. According to undisclosed sources cited by Reuters, Amazon is working on its own AI model called Olympus, with the ambition to compete against OpenAI and Google. This AI model is said to have a massive size, reaching up to 2 million parameters.

Up until now, Amazon has been offering its own AI model called Titan, but it has been relatively small in size compared to its competitors. In order to expand its AI capabilities, Amazon has also invested in Anthropic, a company that provides the Claude AI model on AWS. Additionally, Amazon has formed partnerships with other companies such as A21 Labs and Stability.AI.

The training costs for each LLM model are exorbitant, amounting to hundreds of millions of dollars per iteration. However, Amazon has found various applications for LLM, including powering its voice assistant Alexa, providing product reviews, and even leveraging it to enhance its cloud services on AWS. Therefore, it makes strategic sense for Amazon to invest in training its own AI models.

TLDR: Amazon, in a bid to rival OpenAI and Google, is investing in the development of its own massive AI model called Olympus. The company has already been offering a smaller AI model, Titan, but aims to expand its capabilities with the help of partnerships and investments in companies like Anthropic. Despite the high training costs, Amazon sees the value in training its own AI models for various applications within its ecosystem.

More Reading

Post navigation

Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Surveillance Underway? OpenAI Appoints Former NSA Director as Board Member to Safeguard AI

Astonishing Surge: Arm Stock Skyrockets by a Whopping 48% Post Stellar Performance and Projected Superior Operations, Surpassing Market Expectations

GPT-4o’s Thai Language Tokenizer Test Yields Remarkable Efficiency