Google has agreed to invest $2 billion in Anthropic, an AI startup founded by former executives of OpenAI, the commitment includes an upfront cash injection of $500 million with an additional $1.5 billion for future investments.
Anthropic, established in 2021, is the developer of the artificial intelligence-powered conversational robot Claude 2, a competitor to OpenAI’s ChatGPT robot, which is utilized by companies like Slack, Notion, and Quora.
The company previously received funding from Salesforce and Zoom, valuing it at $4 billion earlier this year.
Claude 2 has the ability to summarize around 75,000 words, equivalent to the length of a book, while ChatGPT can handle approximately 3,000 words. Users can input large datasets and request summaries in various formats such as memos, letters, or stories.
Research by the artificial intelligence company Arthur, which is a platform for monitoring machine learning, found that Claude 2 is the most trusted chatbot in terms of self-awareness, accurately measuring what it knows and doesn’t know, and responding to questions supported only by training data, this conclusion came after Arthur tested chatbots from Meta, Cohere, and OpenAI.
Google previously invested $300 million in Anthropic in April, securing a 10% stake.
Anthropic was one of four companies invited to attend a meeting at the White House in April to discuss the responsible development of artificial intelligence with Vice President Kamala Harris, alongside OpenAI, Alphabet, and Microsoft.
In May, Anthropic raised $450 million in its largest-ever funding round, making it the largest artificial intelligence investment since Microsoft’s investment in OpenAI in January.
Anthropic was founded by Dario Amodei, former VP of Research at OpenAI, and his sister Daniela Amodei, who was OpenAI’s VP of Safety and Policy. Several other graduates of OpenAI research were also part of Anthropic’s founding team.
Amodei stated in July that the company had invested at least two months in developing its latest chatbot, with a team of around 35 people working directly on the artificial intelligence model and a total of 150 supporting it.
