Amazon.com Inc. plans to invest up to US$4 billion in Anthropic, securing a significant partnership in its push to be a leading force in generative artificial intelligence. This move also signifies a strong endorsement of the burgeoning startup.
Under the agreement, Anthropic will primarily utilize Amazon Web Services (AWS) data centres for its software and adopt the cloud giant’s in-house chips for model training, powering chatbots and other tools. This partnership provides Anthropic with both Amazon’s computational capabilities and a substantial financial boost, vital for handling the immense expenses of developing and executing large-scale AI models. The statement released on Monday reveals that Amazon will hold a minority stake in Anthropic. During New York’s premarket trading, Amazon’s stock increased roughly 1%, reaching $130.31.
Historically, Amazon has invested in partners across critical sectors, spanning cargo airlines and grocery distributors to electric truck producers. If the investment in Anthropic approaches the speculated $4 billion, it would be the biggest deal associated directly with AWS. Typically, AWS prefers to develop its solutions rather than procuring external technologies or businesses, especially in a market characterized by soaring, billion-dollar evaluations. However, Amazon stated its engineers, including those beyond AWS, will benefit from Anthropic’s models.
AWS, the globe’s primary provider of on-demand computing power and data storage, has been perceived as trailing in the domain of generative AI, largely because it hasn’t presented a notable product or secured a renowned exclusive ally. Notably, OpenAI, the creator of ChatGPT, relies on Microsoft Corp.’s data centres following a massive $13 billion investment, rejuvenating Amazon’s regional competitor.
Amazon leadership acknowledges the nascent stage of generative AI and has reported that over 100,000 customers have utilized their machine-learning tools. Anthropic’s core model, Claude, was previously accessible via an early-release Amazon service named Bedrock, which offers users both Amazon and third-party models. Amazon CEO Andy Jassy expressed profound admiration for Anthropic’s team and their models, emphasizing the potential benefits of enhanced collaboration for customer experiences.
This partnership is also significant for Amazon’s internal chip production, featuring chips like Trainium and Inferentia developed for machine-learning applications. In contrast to the prevalent AI applications dependent on high-end Nvidia Corp. chips, Anthropic will harness AWS chips for its upcoming foundational models.
Anthropic, established by ex-OpenAI professionals, has already accumulated over $1 billion. They aim to devise a more secure chatbot for summaries, searches, Q&As, and coding tasks. Among its supporters is Alphabet Inc.’s Google, which injected nearly $400 million into Anthropic. It’s worth noting that Google, an AWS competitor, is also engaged in generative AI model development.
Anthropic’s co-founder and CEO, Dario Amodei, believes that amplifying their collaboration with AWS will pave the way for unmatched opportunities for diverse organizations, harmonizing Anthropic’s state-of-the-art, secure AI systems with AWS’s leading cloud tech.