Amazon is the latest to join the raging tech battle that has seen the likes of OpenAI, Google, Microsoft and Nvidia splurge billions on AI.
It announced yesterday that it is investing up to US$4 billion in Anthropic, which created a rival to OpenAI’s ChatGPT called Claude.
Under the agreement, Amazon will provide cloud services for Anthropic’s critical research and development workloads, and Anthropic will provide AWS customers with early access to its foundation models via Amazon Bedrock, a fully managed service that makes foundation models from Amazon and other companies available through an API.
Anthropic will also use AWS Trainium and Inferentia chips, viewed as alternatives to Nvidia’s powerful AI chips, to train and run its future foundation models.
The initial investment made by Amazon on Monday is believed to be around US$1.25 billion, the Wall Street Journal and the Financial Times reported.
“We have tremendous respect for Anthropic’s team and foundation models, and believe we can help improve many customer experiences, short and long-term, through our deeper collaboration,” said Andy Jassy, Amazon chief executive officer.
This AI give and take is one we’ve seen before, noted Forrester analyst Rowan Curran, and which “puts together another powerful pairing of enterprise expertise and advanced LLM capabilities.”
Earlier this year, Microsoft similarly invested US$10 billion in OpenAI and became its primary backer, providing compute and storage for OpenAI’s workloads, with OpenAI powering Microsoft’s services with its foundation models.
However, according to Curran, the Amazon-Anthropic deal “signals that some of the biggest and most technologically advanced companies in the world are being equaled or surpassed by small startups.”
Anthropic, which was founded in 2021 by former research executives at OpenAI, was valued at around US$5 billion in its most recent fundraising round, and that includes backing from Google, which also invested US$300 million in the company last year. Anthropic said at the time that it would train its models on Google’s chips and use its cloud services.
Dario Amodei, co-founder and CEO of Anthropic, clarified in a statement to the Financial Times that “nothing has changed” in its arrangement with Google.
Further, with the strategic collaboration announced today, Amazon zeroed in on AI safety, hoping to set itself apart from the missteps of its competitors.
Anthropic has in fact long touted its products as a safer alternative, claiming that Claude is governed by a set of built in ethical principles that serve as its “constitution.”
Anthropic’s emphasis on the creation and delivery of responsible AI technologies, paired with Amazon’s focus on privacy and security when using their own models in Bedrock, “could potentially be an appealing alternative for companies and governments who want to deploy LLMs at scale in a secure cloud environment,” noted Curran.
Claude, which is available in the U.S. and the U.K., “is much less likely to produce harmful outputs, is easier to converse with, and is more steerable compared to other foundation models,” the companies said in a release.
The collaboration announced on Monday, comes less than a week after Amazon unveiled, at its annual hardware event, future plans for its popular digital assistant, Alexa to be “smarter and more conversational”, powered by generative AI.
This summer, Amazon also spent US$100 million on the AWS Generative AI center, which assembled a team of AI experts to work alongside customers throughout the process of building generative AI systems.