|
|
|
|
August 21, 2025
|
Hackers Infiltrate Alleged North Korean Operative’s Computer, Leak Evidence of...
|
|
August 21, 2025
|
Ecosia Proposes Unusual Stewardship Model for Google Chrome
|
|
August 21, 2025
|
OpenAI Presses Meta for Evidence on Musk’s $97 Billion Takeover Bid
|
|
August 15, 2025
|
ChatGPT Mobile App Surpasses $2 Billion in Consumer Spending, Dominating Rivals
|
|
|
Meta’s Superintelligence Lab Considers Pivot from Open Source AI Model Behemoth to Closed Systems
July 14, 2025
Top members of Meta’s new Superintelligence Lab are reportedly discussing a major strategic shift: moving away from the company’s powerful open source AI model, Behemoth, and instead focusing on developing a closed-source model, according to The New York Times.
Sources told the outlet that Meta had completed training Behemoth but delayed its public release due to disappointing internal performance results. Following the launch of the Superintelligence Lab, testing on Behemoth was reportedly paused, sparking talks about deprioritizing the open source approach.
That said, these discussions remain preliminary. Any significant change would require approval from Meta CEO Mark Zuckerberg. A company spokesperson told TechCrunch that Meta’s stance on open source AI remains “unchanged.”
“We plan to continue releasing leading open source models,” the spokesperson said. “We haven’t released everything we’ve developed historically and we expect to continue training a mix of open and closed models going forward.”
However, the spokesperson declined to comment directly on the potential shift away from Behemoth. If Meta does decide to prioritize closed-source models, it would mark a major philosophical departure from its previous commitments.
Meta currently uses more advanced closed-source models internally, such as those powering its Meta AI assistant. Still, Zuckerberg had positioned open source as a cornerstone of the company’s external AI strategy, viewing the openness of the Llama family as a key differentiator from rivals like OpenAI — which Zuckerberg publicly criticized for becoming more closed after its partnership with Microsoft.
The reality, though, is that Meta faces mounting pressure to monetize beyond its traditional advertising business. The company is investing billions in AI, including paying massive signing bonuses and nine-figure salaries to lure top talent, building new data centers, and funding the costly pursuit of artificial general intelligence (AGI), or “superintelligence.”
Despite boasting one of the world’s top AI research labs, Meta trails competitors like OpenAI, Anthropic, Google DeepMind, and xAI in turning research into profitable products.
A shift toward closed models could suggest that Meta’s openness to date was more strategic than ideological. Zuckerberg’s past remarks hint at this ambivalence. On a podcast last summer, he said:
“We’re obviously very pro open source, but I haven’t committed to releasing every single thing that we do. I’m basically very inclined to think that open sourcing is going to be good for the community and also good for us because we’ll benefit from the innovations. 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. It’s all very difficult to predict.”
Closed models would give Meta greater control and more monetization options—especially if it believes its newly acquired talent can deliver top-tier performance.
Such a pivot could reshape the AI landscape. The momentum behind open source AI, largely propelled by Meta and models like Llama, might slow, even as OpenAI prepares to release its long-awaited open model. Power could consolidate back to major players with closed ecosystems, while open source development becomes driven primarily by grassroots communities.
This shift would ripple through the startup ecosystem, particularly affecting smaller companies specializing in fine-tuning, AI safety, and model alignment that depend on access to open foundation models.
On the global stage, Meta’s potential retreat from open source could cede influence to China, where open source AI projects like DeepSeek and Moonshot AI are embraced as tools for advancing domestic capabilities and expanding global reach.
|
|
|
Sign Up to
Our Newsletter!
Get the latest news in tech.
|
|
|
