Meta is reportedly preparing to establish a new research lab dedicated to the development of artificial “superintelligence,” a form of AI theorized to be vastly more intelligent than humans, signaling its heightened ambitions in the competitive AI landscape.
The new lab aims to push the boundaries of AI capabilities beyond artificial general intelligence (AGI), defined by some as a theoretical computer capable of outperforming humans across all economically valuable tasks. Superintelligence is conceived as an intellect so advanced that its full capabilities are beyond human comprehension.
Alexander Wang, the 28-year-old founder of Scale AI, a startup specializing in building AI applications, is reportedly set to join the new Meta lab. This potential move comes as Meta is also rumored to be in discussions to invest a significant sum in Scale AI, with Bloomberg reporting that this investment could exceed $10 billion.
This strategic maneuver is the latest development in the ongoing global race for AI supremacy, which gained significant momentum following the mainstream launch of ChatGPT in 2022. Meta has already integrated its Meta AI chatbot product across its social media platforms, including Facebook, Instagram, and WhatsApp, and into its smart glasses.
Additionally, Meta has adopted a strategy of open-sourcing its AI systems, including its Llama family of large language models, positioning itself as a more developer-friendly alternative compared to some of its competitors.
The pursuit of advanced AI by major tech companies like Meta reflects a significant industry-wide shift, increasingly relying on the acquisition of newer, AI-focused startups. For instance, Google acquired DeepMind in 2014, later merging it with Google Brain to form Google DeepMind. Microsoft has been making substantial investments in OpenAI, while Amazon has been similarly investing in Anthropic.
Meta’s potential multi-billion dollar investment in Scale AI is expected to bolster its AI development efforts, particularly at a time of escalating competition for limited computing resources and top AI talent. The Times report, citing anonymous sources, also noted that Meta is offering significantly generous salaries, reportedly ranging from seven to ten figures, to attract researchers from rival firms such as OpenAI and Google.
The term “superintelligence” was popularized by Oxford philosopher Nick Bostrom in his 2014 book of the same name. Bostrom’s work included warnings about the potential dangers of an AI “intelligence explosion,” where computing capabilities advance exponentially and potentially evade human control.
While some now view Bostrom’s views as alarmist, concerns about the existential risks posed by an AI system with intellect far surpassing human capabilities persist among some in Silicon Valley, sometimes referred to as “doomers.” However, the prevailing capitalist drive to build and innovate within the AI space, often driven by those dubbed “boomers,” has largely overshadowed these fears thus far.
