Andy Jassy said in his annual letter to shareholders Thursday that the Seattle-headquartered e-commerce company has been developing its own large language models “for a while” and believes the technology will “transform and improve virtually every customer experience.” (The Logic)

Talking point: Amazon already uses machine learning to make personalized e-commerce recommendations and optimize fulfillment centre pick-up paths, for example. The web giant began delivering its first LLM training chips last year, which it claims are faster and cheaper than GPUs. The company will also make generative AI technology available to other firms through its cloud services unit, Amazon Web Services. This comes as Big Tech firms jostle to pull ahead in the AI race. Earlier this year, Microsoft invested US$10 billion into Open AI, the San Francisco-firm behind the wildly popular ChatGPT, and Google announced the release of rival product Bard. Jassy’s letter also warned of short-term headwinds for AWS as companies spend less under the current economic climate.