Google will invest up to $US2 billion in Anthropic, a competitor to OpenAI, ramping up the race between artificial intelligence companies, originally reported by The Wall Street Journal Friday. Amazon committed a $US4 billion investment to the startup and its AI assistant Claude last month.
Gizmodo confirmed the deal with an Anthropic spokesperson Monday. Former OpenAI executives founded Anthropic after a dispute with CEO Sam Altman around safety measures in AI forced them to leave the company. Google already invested $US55o million, and the WSJ reports the company has a multiyear, $US3 billion deal with Google Cloud.
Google declined to comment further on the deal.
Claude is a powerful AI that can outperform OpenAI’s ChatGPT in certain circumstances. It can summarize up to 75,000 words whereas ChatGPT-4 can only handle about 25,000 words. Anthropic’s AI runs on ‘Constitutional AI,’ which uses a list of ethical guiding rules and principles to create less harmful outputs. The constitution requests that Claude choose responses that shun racism, sexism, and toxicity, among other safeguards.
Amazon is also betting big on the safety-focused generative AI with a $US4 billion investment in Anthropic announced last month, making Amazon Web Services the primary cloud provider for the AI company. The collaboration allowed Anthropic to train its AI foundational models using AWS Trainium and Inferentia chips. In return, AWS gained a small stake in the company that builds Claude, and early access to Anthropic features for AWS customers.
The Biden administration announced sweeping regulation on artificial intelligence companies Monday to foster greater safety, privacy, and trust in the field. Claude is generally considered to be one of the more brand-safe AIs on the market due to its constitution while also being extremely powerful.
The rest of big tech is trying to catch up to Microsoft, which reported a revenue jump of 13% to over $US56 billion in earnings last week, cashing in on its early investment in OpenAI.