Anthropic, OpenAI Team with Private Equity to Rev Up AI Adoption

There’s been a slight snag with the AI rollout: It’s turned into a slow roll.

To speed up adoption, both Anthropic and OpenAI announced joint ventures Monday with private equity titans to help companies integrate AI into their core operations. And their new JV partners have identified the perfect clients for their new product: their own portfolio companies. A captive audience never hurts.

Forward-Deployed Thinking

While Anthropic and OpenAI are eager for every non-tech industry to have a “Claude-pilled” moment of their own, AI firms see slow adoption as a deployment problem. The result is the rise of the so-called “forward-deployed engineer,” a techie/consultant/salesperson who pitches businesses on potential avenues for AI implementation. 

OpenAI’s JV, literally called The Deployment Company, is entering the world with more than $4 billion in funding at a $10 billion valuation, backed by TPG, Brookfield Asset Management, SoftBank and Bain Capital, according to Bloomberg. Anthropic’s yet-to-be-named JV has scored around $1.5 billion in commitments from the likes of Blackstone, Goldman Sachs, Hellman & Friedman, Sequoia and Singapore’s sovereign wealth fund, per the Financial Times.

The dual push comes as planned massive IPOs put adoption in the forefront:

• Last week, The Wall Street Journal reported that OpenAI recently missed key revenue and user targets, prompting CFO Sarah Friar to warn that the company may not be able to afford looming compute power contracts. The firm has been steadily losing ground to Anthropic, sources told the WSJ.
• The Deployment Company will open the doors to some 2,000 portfolio companies and clients held by its private equity partners, sources told Bloomberg. 

Private equity, whose portfolios are loaded with software firms vulnerable to an AI wipeout, has plenty of ulterior motives, too. “The near-term value to our portfolio companies is substantial,” Hellman & Friedman CEO Patrick Healy told the FT.

Crunch Time: Boosting adoption may force the industry to reckon with its next great problem: compute supply. Anthropic’s Claude model has recently faced widespread allegations of performance decline after the company reduced its default “effort” level, likely due to a power crunch, and Google CEO Sundar Pichai even met with US government officials last week to discuss the capacity shortage. Meanwhile, Project Stargate, OpenAI’s $500 billion plan to build data centers, has all but petered out, according to the FT. The result is the new catch-22 for AI firms: too much adoption for its current level of compute supply, not enough adoption to fund more compute supply.