Google invested $300 million in AI firm founded by former OpenAI researchers

The competition between Google and Microsoft over tech’s AI future continues to heat up. While Microsoft is firmly entangled with ChatGPT creator OpenAI, Google may have turned to a company founded by former OpenAI employees: the little-known Anthropic.

According to a report from the Financial Times, in late 2022, Google invested around $300 million in the startup, though the news was not reported at the time. In return for the money, Google got a 10 percent stake in the company, and Anthropic is required to buy cloud computing resources from the search giant, says the FT.

This dynamic is somewhat similar to the partnership between Microsoft and OpenAI. There, OpenAI provides the research expertise, while Microsoft supplies not only billions of dollars in investment but also access to its mammoth cloud platform, which is needed to train the latest compute-intensive AI models.

It’s notable that Anthropic is also developing its own general-purpose chatbot: a potential rival to ChatGPT named Claude. (Claude is in closed beta, but you can read a comparison between Anthropic and OpenAI’s systems here.) However, it’s not clear that Google plans to integrate Claude into its services as Microsoft is doing with ChatGPT. Google already has plenty of in-house expertise developing AI language systems, and FT suggests one motivation for the deal is simply building up Google’s cloud computing business.

The back story for Anthropic’s founding is relevant, too. The organization was created in 2021 as a public benefit corporation by Dario Amodei, a former vice president of research at OpenAI. Amodei took with him a number of researchers from OpenAI, including Tom Brown, lead engineer for AI language model GPT-3. As the FT notes, Amodei split from OpenAI “after a disagreement over the company’s direction” — namely, the company’s increasingly commercial focus after its first deal with Microsoft in 2019.

Many AI researchers have since criticized the actions of OpenAI as increasingly reckless, particularly its launch of ChatGPT onto the public web late last year without proper safeguards or software that could reliably detect its output. In comparison, Anthropic emphasizes its work building “reliable, interpretable, and steerable AI systems” on its website. But will Google’s investment affect a shift in these priorities?

At any rate, we can expect to hear more about Google’s AI future very soon. The company is holding an event on the topic next week, February 8th. And we’ll likely be seeing more from Microsoft in the near future about its plans to add ChatGPT to Bing.