28-year-old MIT dropout who is fighting OpenAI with Scale AI

Meta Platforms is acquiring a 49% stake in data-labeling company Scale AI for $15 billion, valuing the company at $29 billion. The transaction, announced by both companies on Thursday, is a strategic investment for Meta as it desperately tries to catch up with artificial intelligence.

Scale AI CEO and MIT dropout Alexandr Wang, aged 28, will travel to Meta to head up its newly-established superintelligence team. This department creates systems beyond current levels of artificial intelligence—to artificial superintelligence (ASI).

"We'll expand the work we do together generating data for AI models and Alexandr Wang will come to Meta to assist our superintelligence efforts," Meta stated in a release, according to Reuters.

Though Meta won't get a board seat at Scale AI, the deal will involve some of Scale's 1,500 employees come to Meta under Wang. Wang will remain a board member at Scale.

Who is Alexandr Wang? The man Meta is putting its hopes on

Wang's own background is unconventional. Born and raised in Los Alamos, New Mexico, by Chinese immigrant physicists, Wang entered the technology world at a young age. He once worked at Quora prior to dropping out of MIT in his first year. In 2016, together with Lucy Guo, he co-founded Scale AI via startup accelerator Y Combinator.

"Long-term, we want to power any human-process-driven business for any company," Wang said to the YC blog in 2016.

Only then 24 years old and the world's youngest self-made billionaire, he was just beginning. Guo left the startup a couple of years later, but Wang founded Scale AI as a data stack for much of the world's top AI systems.

He's raised more than $680 million, including $100 million from Founders Fund co-founder Peter Thiel. Forbes now estimates his net worth at $3.6 billion.

"Focus on making the business and then all the other stuff will just kind of work itself out," he said to Business Insider in 2020.

Wang has become a familiar presence in Washington, frequently meeting with lawmakers about the national security implications of AI. A 2018 trip to China convinced him that America's future warfare would be dependent upon AI leadership.

"AI leadership globally is already in full swing, and the ability of our country to adopt and leverage AI effectively will shape the future of warfare," Wang testified at a public hearing.

Scale AI, established in 2016, assists in training edge AI models by providing large volumes of labeled data. Remotasks and Outlier, its two platforms, employ gig workers to label giant datasets. Labeled data of this nature is essential to train AI models such as ChatGPT.

It started by backing autonomous vehicle customers like Toyota, Honda, and Waymo. It has added backing for OpenAI, Microsoft, and even the US government, which uses it to process satellite imagery from Ukraine.

Scale's 2024 revenues came in at $870 million and will double more than double to reach $2 billion in 2025. Bloomberg says this would value it at $25 billion.

However, the meteoric rise of the startup hasn't all been smooth sailing. There have been reports of substandard working conditions for its offshore gig workers who earn as little as $1 per hour. They are predominantly based in nations such as Kenya, the Philippines, and India.

Meta's AI gamble: Wang bucks research wisdom

This is not just an investment—it's a statement. By doing this, Meta is making the statement that it is giving up on the antiquated research-driven strategy that it used to promote.

Internal battles, such as high-profile departures of talent and model-latency issues, have sapped Meta's AI momentum. Its open-source LLaMA models were meant to shake up the sector, but poor liftoff and talent flight have silenced momentum.

Meta's seasoned AI head Yann LeCun is back in charge. But his skepticism regarding large language models (LLMs) as a paBy hiring Wang— who scaled Scale up to a billion-dollar company without research credentials—CEO Mark Zuckerberg is now wagering on another type of leadership. A business brain like Sam Altman's, and not that of a research purist.

Meta is said to be poaching Google and OpenAI employees with seven- to nine-figure offers to man its 50-person superintelligence lab.

"This was a deeply unique moment": Wang takes new role

In an email to employees, Wang acknowledged the emotional significance of departing Scale.

"Being not a Scalien was, in short, unthinkable. But the more time I had to actually think about it, the more I came to see that this was a very idiosyncratic moment for me, but indeed for Scale," he wrote.

He promised Scale's employees that proceeds of the investment by Meta would go to shareholders and vested equity holders.

Wang will oversee a bold initiative at Meta: creating AI that not only catches up to its competitors, but surpasses them. Superintelligence is still in theory—but with Wang at the helm of the project, Meta is making a $15 billion bet that it can become a reality.to artificial general intelligence (AGI) seems to have gone out of step with traditional Silicon Valley dogma.