Google is investing $300M in an OpenAI challenger that will take on ChatGPT while focusing on A.I. safety

0
167
google-is-investing-$300m-in-an-openai-challenger-that-will-take-on-chatgpt-while-focusing-on-ai.-safety

On Friday, Google announced a partnership with Anthropic, a startup founded in 2021. Anthropic makes Claude, which like ChatGPT is an artificial intelligence chatbot. 

With the $300 million deal, Google will take a roughly 10% stake in the startup, according to the Financial Times. Anthropic for its part gets both a financial boost and cloud computing resources it needs. The deal could give Anthropic a valuation of roughly $5 billion, the New York Times reported.

Of course, ChatGPT maker OpenAI has Microsoft money behind it—lots of it. In a deal revealed last month, Microsoft will invest $10 billion in the startup. It also invested $1 billion in 2019, then quietly added another $2 billion in 2021. The software giant plans to incorporate OpenAI technology into a wide variety of products, including the Bing search engine, possibly disrupting Google’s search dominance.

The two startups share some DNA. Anthropic was formed mostly from a group that broke away from OpenAI after growing disillusioned with strategic and cultural shifts. One former OpenAI employee recently told Fortune that while OpenAI founders Sam Altman and Greg Brockman paid lip service to A.I. safety, “that often seemed like just a fig leaf for business concerns, while actual, legitimate A.I. safety concerns were brushed aside.” 

Anthropic says it’s focused on A.I. safety—it calls itself “an A.I. safety and research company”—and points to its research on “Constitutional A.I.”

With the latter, it writes on its website, “We show that language models can learn to follow a set of simple, natural language principles via self-improvement, and we use this new method to train a more harmless assistant.” 

Tesla CEO Elon Musk has warned about the dangers of advanced A.I. In December he called ChatGPT “scary good,” adding, “We are not far from dangerously strong AI.” He also tweeted that his confidence in OpenAI’s safety was “not high,” noting that, with help from him, it “started as open-source & non-profit. Neither are still true.”

Fortune reached out to OpenAI for comment but did not receive an immediate reply.

Anthropic released a limited test of Claude in January and plans to share it with a wider audience. 

“We’re partnering with Google Cloud to support the next phase of Anthropic, where we’re going to deploy our AI systems to a larger set of people,” Anthropic CEO Dario Amodei said in a statement. “This partnership gives us the cloud infrastructure performance and scale we need.”

He added, “We are eager to use the Google Cloud infrastructure to build reliable, interpretable, and steerable A.I. systems.”

Learn how to navigate and strengthen trust in your business with The Trust Factor, a weekly newsletter examining what leaders need to succeed. Sign up here.