Officials at Meta’s Superintelligence Lab are contemplating a move from Meta’s promised open-source AI model, Behemoth, to a closed model, according to The New York Times. Despite completing training on Behemoth, its launch was delayed due to disappointing internal feedback, and testing has halted since the Lab’s formation. These discussions are still in early stages; and, any changes would need CEO Mark Zuckerberg’s endorsement, who has previously affirmed the company’s dedication to open-source AI.
While a representative expressed plans to keep releasing open-source models, she did not mention the potential shift away from Behemoth, which would represent a profound change in Meta’s philosophy. Though Meta boasts an advanced AI research lab, it lags in successfully monetizing its AI efforts. Prioritizing closed models might suggest that its earlier commitment to openness was more of a whim than a strategy. Such a transition could have an impact on the AI sector, reducing the momentum of other open-source initiatives driven significantly by Meta and potentially strengthening China’s global initiatives to bolster its open-source AI capabilities.
The ainewsarticles.com article you just read is a brief synopsis; the original article can be found here: Read the Full Article…
