Meta's $15B Superintelligence AI Investment

According to various reports, Meta is set to announce a $15 billion investment to pursue computerized “superintelligence.” This announcement comes as the competition to dominate the artificial intelligence field intensifies in Silicon Valley, despite the inconsistent performance of current AI systems.

Mark Zuckerberg, Meta’s CEO, is reportedly going to announce that the company will acquire a 49% stake in Scale AI, a startup founded by Alexandr Wang and co-founded by Lucy Guo. One Silicon Valley analyst characterized the move as something “a wartime CEO” would do.

Superintelligence is defined as AI that outperforms humans in every task. Currently, AI has not reached human-level capabilities across all tasks, which is known as artificial general intelligence (AGI). Studies indicate that many typical systems fail when faced with very complex problems.

Observers believe Meta's attempt to bypass the current level of progress and aim for superintelligence is a way for the company to take back the lead in AI. This follows advancements made by competitors like Sam Altman’s OpenAI and Google, and after Meta’s significant investment in the Metaverse underperformed.

Wang signed an agreement in March to develop ThunderForge, a US defense department system that uses AI for US military planning and operations, initially focusing on the Indo-Pacific and European commands. Peter Thiel’s Founders Fund was an early investor in the company.

Meta’s move has spurred calls for European governments to start their own more transparent research initiatives. The goal is to ensure the responsible development of the technology while maintaining public trust, creating an AI equivalent to the Cern European Organization for Nuclear Research in Switzerland.

Michael Wooldridge, a professor at the University of Oxford, suggested that governments should collaborate to openly and robustly develop AI to ensure understanding and trust in the technology. He stated that the reported deal “seems very much like an attempt by Meta to regain the initiative after the Metaverse didn’t take off and was ridiculed.” He added that AI development is uneven and that AGI is still a long way off, let alone “superintelligence”, saying that while AI can do impressive things, it can also fail at simple tasks.

Dr. Andrew Rogoyski from the Institute for People-Centred AI at Surrey University, stated that the deal shows that AI companies are acquiring AI talent from various sources, including young AI companies and university AI departments. He noted that Meta’s AI strategy differs from companies like OpenAI and Anthropic, as AI is used to support Meta’s business rather than being its primary focus. He said, “It means they’re not quite as desperate to achieve AGI, so they can afford to take a longer view.”

Reports indicate that Wang will take a senior position at Meta. Meta has not provided a comment. Scale AI has been contacted for their response.