Ingo Payments Generation Instant Overpayment Disbursements June 2024 Banner

Amazon Reportedly Developing In-House Answer to ChatGPT

Amazon building

Amazon is reportedly developing its own artificial intelligence (AI) chatbot to compete with OpenAI’s ChatGPT.

The project is code-named “Metis,” Business Insider reported Monday (June 24), noting that the name seems inspired by the Greek goddess of wisdom. 

The chatbot would be accessed through a website and powered by an in-house AI model dubbed Olympus, keeping with the Ancient Greece theme, the report said, citing internal documents and sources familiar with the matter. 

The sources added that Olympus is a more powerful iteration than Titan, the company’s existing and publicly-available AI model. Amazon CEO Andy Jassy is said to be personally involved in the project, the report says.

PYMNTS has contacted Amazon for comment but has not yet gotten a reply.

Business Insider cites an internal document that says Metis provides text-based and image-based answers in a smart, conversational fashion, and can share links to the source of its responses, offer follow-up questions and generate images.

According to the report, sources say Amazon wants Metis to employ an AI technique known as retrieval-augmented generation (RAG), which will generate more up-to-date replies. For instance, Metis should be able to offer up the latest stock prices, something non-RAG chatbots wouldn’t be able to do.

As PYMNTS wrote earlier this year, RAGs are an advancement that could redefine AI by combining information retrieval with natural language generation.

AI large language models (LLMs) have led AI technology and improved various applications, though their tendency to generate false information has stifled their potential. RAG lets AI access and incorporate specific external data into its responses, which makes them more effective and accurate. 

“The main advantage of RAGs over LLMs is the fact that the former is based entirely on a proprietary data set that the owner of said RAG can control, allowing for more targeted applications,” Renat Abyasov, CEO of the AI company Wonderslide, told PYMNTS. 

“Let’s say a doctor wants to deploy a chatbot for their patients; using an RAG will allow them to ensure that the advice provided by said chatbot will be reliable and consistent. That reliability is much harder to achieve with LLMs, systems trained on massive amounts of publicly available and sometimes rather dubious data.”

Amazon’s Jassy, meanwhile, has predicted that AI will help the company generate “billions” in revenue in the coming years.

“Gen AI is and will continue to be an area of pervasive focus and investment across Amazon, primarily because there are few initiatives that give us the chance to reinvent so many of our customer experiences and processes,” he said during an earnings call in March. 

“And we believe it’ll ultimately drive tens of billions of dollars of revenue for Amazon over the next several years.”