In this article META Follow your favorite stocks CREATE FREE ACCOUNT
Scale AI CFO Dennis Cinelli. Courtesy of Scale AI.
After Meta shocked the tech world in June by announcing plans to invest $14.3 billion in Scale AI, primarily as a way to hire the startup's founder, Alexandr Wang, and a handful of his employees, the future of Scale was immediately thrown in doubt. OpenAI soon disclosed that it had been winding down its work with Scale, which prepares the data that artificial intelligence labs and big tech companies use to train their models. Companies including Google and Elon Musk's xAI also paused work with Scale after the deal, according to multiple media reports. But almost five months after the blockbuster announcement, Scale AI CFO Dennis Cinelli has a very different message to share. The 1,000-plus person company, he says, is alive and well. "People mischaracterize this deal as some sort of acquihire or some sort of licensing deal, which is not true," Cinelli, who joined Scale in 2022, told CNBC in an interview. "We're a company that has signed some of the best deals we have had in the history of our company, just in the last two, three months." Founded in 2016, Scale is best known for its data business where it competes with companies including Appen, Surge AI and Mercor. Scale also has an applications business that creates custom solutions to help governments and large enterprises deploy AI. The U.S. Department of Defense signed a $99 million contract with Scale in August, followed by another $100 million contract in September.
watch now
Cinelli, 42, said both parts of the business are growing and bringing in revenue that's "well into the nine figures," though he declined to share more specific revenue numbers. The startup generated close to $1 billion in revenue last year, before the Meta deal, a spokesperson told CNBC in August. Whether Scale has a viable path forward has been a big topic in Silicon Valley. The Meta transaction was lumped in with a number of acquihires that have taken place in the industry, though at a notably higher price. Since early last year, Microsoft , Amazon and Google have each orchestrated deals to bring in top AI talent and license certain technology without having to face the regulatory hassles that come with full-blown acquisitions. In July, Google spent about $2.4 billion to hire Windsurf co-founder and CEO Varun Mohan and other senior research and development employees. The agreement didn't include an investment and came with a nonexclusive license to some of Windsurf's technology.
Already exited?
Greg Martin, who works on pre-initial public offering stock sales as managing director at Rainmaker Securities, called the Meta-Scale arrangement a "quasi-acquihire," and said investors don't know what to make of the company because "it doesn't feel like it can go public anymore." He said he's seeing much more investor interest in names like Anthropic and xAI. "People don't truly understand what the future exit is for Scale," Martin said. "Or did it kind of already have its exit?" Scale has been trying to convince potential investors, employees and customers otherwise. Jason Droege, who was promoted to interim CEO at Scale from strategy chief following Wang's departure, said in a blog post days after the deal was announced that the startup wasn't pivoting or "winding down."
FILE PHOTO: Jason Droege speaks at the WSJTECH live conference in Laguna Beach, California, U.S. October 22, 2019. Mike Blake | Reuters
Cinelli said that Scale continues to work with "all the major AI labs and tech companies," though he declined to comment on specific customers. "We had conversations, everyone had questions, naturally," he said. "We're still an independent company, we're not captive to Meta, we're still in business with all our customers. Once we've gone through those conversations, I think everyone kind of understood." OpenAI, Google and xAI are not currently listed as customers on Scale's website, and representatives from the companies didn't respond to CNBC's request for comment. Meta is named as a client.
Scaling after Wang
... continue reading