Summarize this content to 2000 words in 6 paragraphs
Dario Amodei, CEO of Anthropic, speaks at AWS re:Invent in 2023. (GeekWire File Photo / Todd Bishop)
Amazon will invest another $4 billion in Anthropic, the startup behind the Claude chatbot and AI platform, bringing its total investment to $8 billion under an expanded partnership between the companies.
Even with the additional investment, the Seattle tech giant will remain a minority investor in Anthropic, the San Francisco company said in a post Friday morning. The companies did not disclose Amazon’s percentage stake. The additional investment does not give Amazon a seat on Anthropic’s board of directors or trust.
The move further positions the Amazon-Anthropic combination as a counterpart to the duo of Microsoft and OpenAI. The Redmond-based tech company has invested nearly $14 billion in the ChatGPT maker.
Both partnerships illustrate the symbiotic relationship between cloud giants and leading AI startups, which need access to vast computing and financial resources to train their cutting-edge foundation models.
These types of deals might have been outright acquisitions instead in the past, in a more permissive regulatory environment. Tech industry leaders are watching closely to see what happens to M&A under the Trump administration.
Under their expanded relationship, Amazon Web Services will be Anthropic’s primary partner for training AI models, in addition to its existing role as the AI startup’s primary cloud provider, Amazon said in its post.
Anthropic will train and deploy its future foundation models using AWS Trainium and Inferentia chips, the company added.
That pledge to use Amazon’s chips, rather than Anthropic’s preferred Nvidia processors, was a key deal point for Amazon, according to a Nov. 7 report by The Information. In its post today, Anthropic described its work with Amazon’s Annapurna Labs to optimize AWS Trainium chips for AI model training.
“Through deep technical collaboration, we’re writing low-level kernels that allow us to directly interface with the Trainium silicon, and contributing to the AWS Neuron software stack to strengthen Trainium,” Anthropic said. “Our engineers work closely with Annapurna’s chip design team to extract maximum computational efficiency from the hardware, which we plan to leverage to train our most advanced foundation models.”
Anthropic also noted that Claude’s inclusion in Amazon’s Bedrock generative AI service has made Claude “core infrastructure for tens of thousands of companies.”
In addition, Amazon Web Services and Anthropic will give AWS customers early access to a new capability that allows them to fine-tune their own data on Anthropic models. Amazon described it as “a customization benefit that AWS customers will uniquely enjoy for each model for a period of time on new Claude models.”
Amazon announced its initial investment in Anthropic in September 2023. Google Cloud, one of Amazon’s main cloud rivals, reportedly invested $300 million in Anthropic for a 10% stake as part of a prior funding round.
The announcement comes in advance of the annual AWS re:Invent conference Dec. 2-6. AI has become a key area of competition and potential growth for AWS, Microsoft Azure, Google Cloud and other major cloud providers.
