Tech Updates

Inflection AI, The Year-Old Startup Behind Chatbot Pi, Raises $1.3 billion.

Backed by Microsoft, Nvidia, and billionaires Reid Hoffman, Bill Gates, and Eric Schmidt, the startup led by ex-DeepMind leader Mustafa Suleyman is valued at $4 billion and claims to have the world’s best AI hardware setup.

Less than two months after the launch of their first chatbot, Pi, artificial intelligence startup Inflection AI and CEO Mustafa Suleyman have raised $1.3 billion in new funding.

Microsoft, Nvidia, and three of tech’s most influential billionaires led the investment in the Palo Alto-based startup launched in early 2022. LinkedIn cofounder Reid Hoffman, Microsoft cofounder Bill Gates, and former Google CEO Eric Schmidt all personally invested, with Nvidia the sole new investor among the group.

The new funding values Inflection at $4 billion, according to a source with knowledge of the transaction. Inflection said the company and Suleyman remained majority shareholders and declined further comment.

In an interview, Suleyman said that the group of mostly insiders proposed the additional investment after Inflection was “overwhelmed with offers” following the launch of Pi, its conversational chatbot, in May. “I think people can see that it’s just the tip of the iceberg,” Suleyman told Forbes. “There’s so much further to go after Pi  validates the core thesis, which is that conversation is the new interface.”

Some details of Inflection’s new deal with Microsoft and Nvidia are, like Suleyman’s iceberg, still largely out of view. He declined to provide a breakdown of how much of the $1.3 billion raised included cash equivalents (such as computing credits), but said that a “very, very large chunk” was in dollars. “We have all the cash we need to run and operate,” he added.

Inflection also declined to comment on how much equity Microsoft and Nvidia now hold in the business. But Suleyman said neither company commands ownership-like control over it or other preferential rights. “In practice, it was a very traditional round,” he said. “There’s no IP movement, and we are still entirely independent and at liberty to do whatever we want on the commercial front and partner with whomever we want. So there are no real restrictions,” he said.

What is clear is that the round significantly deepens Inflection’s ties with Microsoft and Nvidia, two key partners in the AI race. Microsoft, also a major investor in OpenAI, is Inflection’s cloud computing partner; Nvidia, meanwhile, has been working closely with Inflection on the deployment of its flagship H100 graphics processing unit (GPU), the current gold standard for AI training and powering large language models like OpenAI’s GPT-3. Nvidia worked closely with Inflection and service provider CoreWeave to co-develop Inflection’s current H100 cluster; Inflection paused its own work for Nvidia to run a recent test that Nvidia announced this week had set records on eight tests of current AI model training benchmarks, completing a benchmark based on GPT-3 in less than 11 minutes.

That test, which matched the computational power of training a model that took an estimated three to six months to develop, ran on Inflection’s 3,584 H100 GPUs already in service, Suleyman noted. But in the wake of this funding round and partnership, Inflection’s growing horsepower is about to get a turbocharge. Nvidia and CoreWeave (which helps physically deploy the GPUs) are now in the process of helping Inflection install many thousands more. Once fully operational, Inflection’s new cluster will run 22,000 H100s.

Inflection believes that it will be the largest GPU cluster for AI applications in the world, ahead of Meta’s 16,000 GPU cluster announced in May. (Just how many OpenAI is using is currently unknown; Nvidia announced last November it planned to incorporate “tens of thousands” of GPUs into Microsoft’s Azure cloud service.) Against the world’s largest clusters overall, Inflection said it estimated that it would trail only Frontier, the supercomputer maintained by the Oak Ridge National Laboratory in Tennessee.

“Microsoft has been amazing; they’re turbo-charging us; they are our anchor,” Suleyman said. “And as a result of our collaboration with Nvidia, we’ve been able to tune our cluster to get it to be the absolute best in the world. We can objectively say now that we have the very best hardware in the world.” Microsoft CTO and AI lead Kevin Scott and Nvidia CEO Jensen Huang declared their support for Inflection in statements, with Huang saying that “the world-class team” at Inflection would help usher in “amazing personal digital assistants.”

With this latest funding, Inflection plans to continue expanding its computing capabilities and further developing Pi, a bot trained to engage in a conversational back-and-forth with users to tease out more valuable questions and answers. That chatbot, which Forbes previewed in early May, is “seeing huge engagement” as it adds new features, according to Suleyman. Inflection declined to provide user numbers to corroborate that claim.

The model used for Pi, which Inflection announced earlier in June and said had similar computing capabilities to OpenAI’s ChatGPT, is one of the smaller ones it’s had in the works, Suleyman claimed, with much bigger to come. Also in the short-term pipeline is an API for selected partners to train their own conversational AIs. (Inflection doesn’t plan to make that generally available as it focuses on its own consumer products, Suleyman said.)

All of that means more funding is imminent—I expect we will continue to raise at an accelerated rate—but don’t expect Inflection to raise it from venture capital firms. “Our network and reach aren’t something that regular VCs can help accelerate,” Suleyman said. “What I’m personally looking for as a CEO is just advice, to benefit from the wisdom of their experience.” (The large sums of cash help, too.)

For skeptics raising their eyebrows at the dollar amounts involved, both for the cost of training large language models and for the rounds their creators are raising, Suleyman offered a couple of explanations. While the cost of training AI models has decreased over time, Inflection and its rivals’ desire to train ever-bigger ones means that their absolute spend continues to grow, he said. Then there’s the “tidal wave” of investor and consumer interest in such technologies—oone that Suleyman is happily “surfing.” “It’s totally nuts,” he admitted. Facing a potentially historic growth opportunity, Suleyman added that Inflection’s best bet is to “blitz-scale” and raise funding voraciously to grow as fast as possible, risks be damned.

“There is, you know, ‘small is beautiful, and we have demonstrated that in under a year,” Suleyman said. “This is the beginning of a journey; really, it’s just the starting line.”



Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button