Chesnot/Getty Images
The most powerful tech leaders in Silicon Valley have been racing to build artificial general intelligence (AGI), a theoretical computer that, as some would define it, can outperform humans across every economically valuable task.
Also: Apple’s Goldilocks approach to AI at WWDC is a winner. Here’s why
Some leaders have their sights set even higher, on artificial “superintelligence,” a system that’s unfathomably more intelligent than human beings, so much so that our comparatively puny intellects can’t even imagine what such a system would be capable of; it’d be like an ant trying to converse with John von Neumann.
Now, Meta is slated to launch a research lab devoted to building superintelligent AI, according to a report from The New York Times. Alexander Wang, the 28-year-old founder of Scale AI, a startup that helps companies build AI apps, which could receive a multibillion-dollar investment from Meta soon, will reportedly join the new lab.
Also: The 7 best AI features announced at Apple’s WWDC that I can’t wait to use
The news marks Meta’s latest effort to stay ahead in the ongoing AI race, which has engulfed much of the tech world since the launch of ChatGPT in 2022 revealed the technology’s capabilities to a mainstream audience.
The company has embedded its chatbot product, Meta AI, across its suite of social media platforms, Facebook, Instagram, and WhatsApp, and in its smart glasses. Meanwhile, Meta has sought to position itself as a more developer-friendly alternative to its competitors by open-sourcing its AI systems, including its family of Llama large language models.
The big fish eat the small
Companies like Meta (previously Facebook), which ascended during the rise of the internet and social media in the early 2000s and 2010s have rushed to embrace AI, widely touted as the technological backbone of the future.
Also: The great AI skills disconnect – and how to fix it
That strategic and economic shift has depended largely on the acquisition of newer, AI-focused startups: Google acquired DeepMind in 2014, later fusing it with its internal AI research lab, Google Brain, to create a new division called Google DeepMind; Microsoft has been pouring billions into OpenAI, and Amazon has been doing the same with Anthropic.
Meta’s planned investment in Scale AI, which, according to Bloomberg, could exceed $10 billion, would likely help the tech giant ratchet up its AI development at a time of growing competition for increasingly scarce computing resources and top AI talent.
Also: AI is paving the way for a new type of organization – a Frontier Firm
Meta is also offering seven-to-ten-figure salaries to researchers from competing firms, such as OpenAI and Google, according to The Times report, which cited anonymous sources with ties to the company.
Doom and boom
The term “superintelligence” was popularized by the Oxford philosopher Nick Bostrom in his 2014 book of the same name. The book served in part as a warning about the potential dangers of an AI “intelligence explosion” — computing capabilities that advance suddenly in an exponential upward curve, escaping human control.
Also: AI leaders must take a tight grip on regulatory, geopolitical, and interpersonal concerns
Though Bostrom’s views are now considered by some to be alarmist, there are many in Silicon Valley today (nicknamed “doomers”) who worry about the potential existential risks of an AI system whose intellect is far greater than our own. But so far, at least, the capitalist incentive to build (led by so-called “boomers”) has overshadowed such fears.
