The Information websiteAccording to sources, Amazon is about to launch a large natural language model code-named "Olympus" with a parameter size of up to 2 trillion sets, which is expected to be launched as early as December this year.
Compared to OpenAI's GPT-4, which has a parameter size of 1 trillion, Amazon plans to launch a large-scale natural language model codenamed "Olympus," which has significantly double the parameter size. It may be unveiled during the upcoming re:Invent 2023 and will be used in its e-commerce services, smart devices such as Echo, and is also expected to be integrated into the AWS cloud platform for use by more developers and enterprises.
Previously, Dave Limp, Amazon's Senior Vice President of Alexa and Devices, mentioned that the Alexa digital assistant service will integrate self-generated artificial intelligence in the future, allowing users to interact with it in a more natural way and help users complete more tasks. At that time, Dave Limp also revealed that Alexa may eventually become a standalone paid service because the underlying operating model requires certain execution costs.
Currently, artificial intelligence has become a key development project for Amazon. CEO Andy Jassy previously stated that artificial intelligence will change many business operation models. Amazon has also recently continued to invest in artificial intelligence application projects. For example, it invested $40 billion in Anthropic, allowing Anthropic to provide large-scale natural language models for integration with AWS services, and allowing AWS service users to use this large-scale natural language model to create various application products.
