(Bloomberg) -- Meta Platforms Inc. Chief Executive Officer Mark Zuckerberg touted his views on artificial intelligence in a 6,500-word essay published Monday, emphasizing his belief that wider access to AI models is key to the industry's future.
"The notion that AI is so dangerous that the only safe path is an extreme concentration of power seems inherently problematic," Zuckerberg wrote, challenging a vision of AI championed by startups like OpenAI and Anthropic PBC. "Historically, hoping that an absolute power will benevolently provide for humanity if sufficiently enlightened has not led to safe or positive outcomes."
Zuckerberg's letter is the latest entry in the broader genre of manifestos from AI executives laying out their visions for the emerging technology. More philosophical than technical, these statements seek to establish their company's place in the field and to make the case for how AI should be deployed to support their particular versions of a techno-optimist future.
Here are five essential takeaways from Zuckerberg's missive:
- Meta Still Cares About Open-Source, Open-Weight AI
For years, Meta has said it was all in favor of "open-source" AI models, industry parlance for making the technology's building blocks publicly available for download so that developers can use and modify them. The company invested heavily in its Llama family of models, which first debuted in 2023 but ultimately underwhelmed Silicon Valley. Over the past year or so, though, Meta has pivoted. It hired a new team of researchers, formed Meta Superintelligence Labs and started to prioritize a new generation of models that it could keep proprietary and charge users for access.
While Llama appears to be heading toward retirement, it's clear from Zuckerberg's letter that the CEO hasn't entirely abandoned the idea of giving developers broader access to Meta's AI products. In his essay, he refers to open source at least 16 times, stressing the importance of US leadership in this arena. On Monday, the company announced a new open-weight system called Muse Glimmer. Users can download and alter the model, but won't have access to all the ingredients that went into making it. Meta said it intends to make the weights for a version of its more powerful Muse Spark available as well. Those weights refer to the internal values that the model learned during training and that help determine how it responds to future prompts.
- Concentration of AI Control Poses a Major Risk
Zuckerberg also raised concerns about the potential for control over AI to be concentrated in the hands of a few large institutions — like businesses and governments — though he stopped short of naming any of Meta's rivals.
"That will naturally lead to outcomes that are less favorable for everyone else," he wrote. "This is not a technological principle. It is about the balance of power. There is no such thing as a singular benevolent superintelligence."
Zuckerberg argued that broad access to AI diffuses power, leading to greater competition. It's a pitch being made by a growing number of tech executives right now — a quiet jab at OpenAI and Anthropic, which have emerged as the clear leaders in the fast-moving race and have been more focused on controlling access to their technology.
- AI Could Expand The Job Market
Zuckerberg rejected concerns aired by policymakers in Washington and even some in Silicon Valley that artificial intelligence will result in "fewer jobs overall."
Instead, the Meta chief sees a world where there are "a larger number of companies with fewer people" employed at each firm. The technology will enable individuals to be more entrepreneurial, he wrote, and each person will be equipped with a "personalized tutor and coach with a Ph.D. in every subject" to help them get ahead.
"People also continually come up with new ideas to make our lives better and new jobs to bring those ideas to life," he wrote. "In the near future, there will be new jobs that aren't common today."
There is no small amount of irony to this take: Zuckerberg cut 8,000 jobs from Meta earlier this year in response to the company's pivot toward AI.
- Data Center Communities Get $1 Billion
Meta on Monday also announced a $1 billion fund to invest in communities where the company owns and operates data centers. One such community is Richland Parish, Louisiana, where Meta is building a data center that may ultimately cost more than $250 billion.
With its pledge, Meta is seeking to address growing concerns over the local impact of new AI infrastructure. Data centers have drawn the ire of communities across the US over fears the facilities will strain nearby resources and drive up water and energy costs.
"Sustainable infrastructure development means that communities must benefit significantly from each project," Zuckerberg wrote, adding that this includes high-paying jobs, and investments into schools and public services. He added that in Richland Parish, local teachers received a $50,000 bonus thanks to increased tax revenue from the Meta investment. However, the school board has been preparing teachers for that bonus to decrease in coming years as the initial benefits from the boomtown wane.
- The Threat From China
Zuckerberg has warned for years that China poses a technological threat to the US. His essay once again outlined what he sees as the risk of losing ground to China if US lawmakers are too heavy-handed with AI rules and regulations.
"American communities will have greater prosperity and security if America and its allies lead in AI compared to geopolitical rivals," he wrote. This is made more challenging because "it is more difficult to build infrastructure here" than in China, he added.
The US will need to speed up its AI development, including its infrastructure projects, to keep pace, Zuckerberg wrote. He praised US measures limiting sales of advanced AI chips and other hardware to China. "Export controls on silicon have been successful for slowing the progress of foreign labs during this critical period, so it is the right strategic move to continue those," he added.