AI data labelling start-up Scale valuation doubles to $14bn

Stay informed with free updates

Scale AI has raised $1bn in a new funding round, doubling the data-labelling start-up’s valuation to about $14bn and paving the way for an initial public offering.

The deal underscores the critical nature of data in the race to build super-powerful artificial intelligence systems and the premium investors are willing to put on a commodity many regard as the new oil.

The San Francisco-headquartered company, which provides accurately labelled data for companies building AI models, has tapped existing investors, including venture capital firms Accel, which is leading the round, Nvidia, Josh Kushner’s Thrive Capital, Tiger Global Management and Index Ventures. New investors include Amazon, Meta, Intel Capital and AMD Ventures.

The round is the latest in a series of AI megadeals. OpenAI and Anthropic have raised billions of dollars in the past 12 months from Microsoft, Amazon and Google, and Elon Musk is in the process of raising as much as $6bn from investors for his own venture, xAI.

But unlike its biggest customers — which include OpenAI, Meta and Microsoft — Scale is not building its own generative AI models.

It was launched in 2016 to label the images used to develop autonomous driving systems. Since then it has grown rapidly by providing enormous volumes of accurately labelled data to train tools such as OpenAI’s ChatGPT.

The company, which was valued at $7.3bn in 2021, employs a vast network of employees, many of them contractors, to do the labelling. Scale’s revenues were roughly $700mn last year, according to a person with direct knowledge of the matter.

Its focus on what Alexandr Wang, co-founder and chief executive, describes as “one of the least sexy problems in AI” has given Scale a central position in the booming sector.

AI models have improved dramatically over the past 18 months, but further leaps — such as the capability to reason, interpret text, images and speech simultaneously, or complete multi-step tasks — rely on larger, more complex data sets.

“As an AI community we’ve exhausted all of the easy data, the internet data, and now we need to move on to more complex data,” said Wang. “The quantity matters but the quality is paramount. We’re no longer in a paradigm where more comments off Reddit are going to magically improve the models.”

One venture capitalist who chose not to participate in Scale’s latest funding round said he was concerned that the company’s revenues were too concentrated with a small clutch of AI businesses that are burning through cash in a fierce competition to build the best models.

“We’re an infrastructure provider,” said Wang. “How the model development ecosystem develops above that obviously affects our business, as it does Nvidia, data centre providers and cloud providers.”

Wang has been courting the US government as well as top AI companies. The 27-year-old has become a prominent Silicon Valley figure in Washington, speaking to lawmakers about the importance of beating China in what he has described as a “prolonged tech cold war between the two greatest economic powers in the world”.

Earlier this year, Scale was awarded a contract by the Department of Defense to test and evaluate various AI models.

The funding round is likely to be Scale’s last as a private company, according to one investor in the business. “The next logical thing is a public debut and there are investors (in the round) who could help with that,” they said.

“As with every private company we always aim for every round to be the last round of funding that we need,” said Wang. “In terms of an IPO, we’re always eyeing the market.”

Source

Leave a Comment