Meta Platforms Inc. has kicked off a multi-year odyssey to develop sophisticated new generative artificial intelligence. They have already staked billions into talent and technology to supercharge this effort. The company has been intent on beating the rest of the field, most notably OpenAI. It is now laser-focused on bringing its next generation of AI—Llama 3—to fruition. Whether or not this means more open-source strategies are in the pipeline for Meta’s AI is anyone’s guess.
In a significant move to bolster its AI capabilities, Meta has acquired top researchers and engineers from leading AI firms and committed to building new data centers. This monumental move makes it very clear how serious Meta is about deepening its artificial intelligence expertise. As I write this, the company’s recently installed executives are laser-focused on creating their opening offering to Llama 3. They are motivated to outdo OpenAI’s GPT-4 model.
In June 2025, Meta made a pivotal investment of $14.3 billion in Scale AI, a notable player in the AI landscape. The acquisition included bringing on board Scale AI’s founder and CEO, Alexandr Wang, who is expected to play a vital role in steering Meta’s AI initiatives. To focus its efforts, Meta has consolidated its AI projects under a new division, Meta Superintelligence Labs.
Mark Zuckerberg, the CEO of Meta, has articulated a broader mission: to introduce “personal superintelligence” to consumers worldwide through innovative products such as augmented reality glasses and virtual reality headsets. Zuckerberg’s right that the personal device is becoming the primary computing platform. These devices will finally start to anticipate context, communicate fluidly, and move with us through our environments.
“Personal devices like glasses that understand our context because they can see what we see, hear what we hear, and interact with us throughout the day will become our primary computing devices.” – Mark Zuckerberg
As of late, Zuckerberg himself has touted Meta’s Llama family of open models as a major differentiator from its competition. He expressed optimism regarding the future of these models in a 2024 letter, stating, “Starting next year, we expect future Llama models to become the most advanced in the industry.” This grand ambition is indicative of Delaware’s desire to be an innovation and industry leader in the rapidly expanding AI space.
It now seems like there may be some wiggle room on Meta’s open-source commitments. He indicated that if there is a significant qualitative change in AI capabilities, the company might reconsider its approach to open-sourcing its models.
“If at some point however there’s some qualitative change in what the thing is capable of, and we feel like it’s not responsible to open source it, then we won’t.” – Mark Zuckerberg
This kind of sentiment is part of why many people are so wary of discussions about the long term impact of superintelligent AI. Although Zuckerberg seems to support the idea of sharing their progress far and wide, he’s clearly aware of the safety issues that come along with it.
“We believe the benefits of superintelligence should be shared with the world as broadly as possible,” – Mark Zuckerberg
Yet he adds, “That said, superintelligence will raise novel safety concerns. We’ll need to be rigorous about mitigating these risks and careful about what we choose to open source.”
Regardless of all these unknowns, when it comes to open-source AI, Meta doubled down. A spokesperson for the company stated, “Our position on open source AI is unchanged. We plan to continue releasing leading open source models. We haven’t released everything we’ve developed historically and we expect to continue training a mix of open and closed models going forward.”
This hybrid of digital and in-person displays a practical hand as Meta continues to maneuver through the quickly changing world of artificial intelligence. The tension between risk and responsibility always has, and still is, the focus of their innovation development process.
