Meta invests $ 14 billion in scale AI to ‘deepen’ his work on super intelligence

The American multinational technology company Meta has made an important investment in scale AI, the Data-Label Startup that is essential for training artificial intelligence systems, recruiting his young founder Alexandr Wang to build a “super intelligence” lab aimed at achieving artificial general intelligence.

The investment of $ 14.3 billion acquired an interest of 49% in scale AI, which increased its appreciation to more than $ 29 billion. This is the second largest investment of Meta, after the WhatsApp acquisition of $ 19 billion in 2014.

Scale AI announced that Wang would continue to serve as a director in the company’s board of directors, with the company’s main strategy officer, Jason Debe, who stepped into his shoes as Interim -CEO.

In one tweetWang shared a memorandum for employees, in which the investment of Meta was described as a “big milestone and a powerful validation of the hard work that you have all brought into the mission of scale.”

For Meta, who has worked on his own AI strategies and recently released a new model and an independent AI app in recent months, the deal “to deepen the work” in which it is involved in “producing data for AI models”, according to business declarations that are shared with media.

Meta’s AI Race

This speaks to the urgency of Meta’s movement to catch up with the AI ​​race, because it is confronted with increasing pressure from other players working on “frontier models” that are advanced, large-scale systems that push the boundaries of AI.

Nevertheless, the relocation also emphasizes an important tension between large technology companies such as Meta and others who build on decentralized platforms, according to Renz Chong, CEO of A16Z-Supported Modular On-Chain Platform Sovrun.

“Earlier this year we saw the rise of open-source frontier models that can go to toe with closed models of Big Tech,” Chong said Decrypt. “That is a clear signal:” state-of-the-art “no longer needs to be centralized or meaning.”

Because most other AI-to-chain projects are still tend to “centralized inference end points or APIs outside the chain,” the perilous situation explains the pressure on decentralized players.

“Early infrastructure players put critical basic work and offer decentralized calculation and stimulated training strokes,” Chong noted.

What is Scale AI?

Scale AI specializes in Databenetings Services that are essential for training AI systems, together with customers such as Google and OpenAi. The startup uses human annotators to classify data that feeds AI models, with much of the work that is carried out outside the US through labor-intensive processes.

The investment of Meta is said to give a minority interest in scale AI, so that the startup can maintain operational independence – a structure that can help Meta to avoid extra regulatory research in the midst of current antitrust fights. Earlier in April, Meta was accused of helping the AI ​​ambitions of China, which encouraged a study by the Senate.

Nevertheless, the consolidation of AI resources between technical giants has given rise to alternative visions from the decentralized sector. Sovrun, for example, recently introduced a joint venture with Virtuals Protocol to build Readygamer, a platform that integrates AI-driven NPCs into popular game worlds.

Although those projects had a decrease in income earlier this year, a revival is underway, with daily figures that slowly return to earlier levels, according to data maintained by the virtual protocol.

Chong argues that the real shift can not only be found to make better systems, but also in “change who they form” and build up “results that are important for the communities they serve.”