Meta is pouring millions of dollars into the new Meta Superintelligence Labs (MSL) to gain an edge in the artificial intelligence race. Meta boss Mark Zuckerberg hired top minds in the AI space to build sophisticated models and achieve artificial general information (AGI). However, on the road to AI domination, Meta reportedly struggles with its own internal systems. As a result, the company is now asking its employees to look to external tools to speed up AI development.
Over the past few years, Meta has made AI the number one priority, with Zuckerberg pouring hundreds of millions of dollars into the field, recruiting leading researchers in a nine-figure pay package, and reshaping AI efforts under MSL. But while the ambition is massive, Business Insider reports that the company's existing internal infrastructure, designed for billions of users and a vast engineering team, is becoming a stumbling block for smaller, faster AI groups.
Citing internal documents, the report reveals that Meta's Product and Applied Research (PAR) group led by former Github CEO Nat Friedman is urging staff to migrate from meta's slower internal systems to widely used developer platforms such as Vercel and Github.
In a memo sent in late September, MSL's Infrastructure Director Aparna Ramani reportedly admitted that Meta's internal tools “take too long” to deploy changes and “doesn't encourage coding.” Vibe coding, where engineers rely on AI to generate and refine code, is gaining momentum across the high-tech industry. But in meta, even small changes can take hours to unfold, compared to just a few minutes on external platforms.
According to the report, Ramani's memo outlines plans to reduce deployment times from 99 to under the age of 2. She explained two strategies: Immediately adopt Vercel to unlock MSL progress, while also developing a nested internal alternative codename optimized to host apps built with TypeScript.
Meanwhile, the documentation shows that Meta has already started pairing vercel with Github, Microsoft's code hosting and collaboration platform, and has begun pairing vercels to provide a faster way to build and test AI applications. By mid-September, at least 10 active projects were already running in Vercel, and updates were rolled out in minutes rather than hours.
Vercel is a cloud platform where businesses build and deploy web applications, focusing on front-end frameworks and full-stack apps. Provides tools, infrastructure and global networks to provide a high-performance, scalable, and secure web experience. Interestingly, both Meta's new top AI executives, Friedman and Alexandre Wang, are Versel investors. The company has also become one of the most prominent names in web and AI app development, raising $300 million recently at a $300 million valuation. Its customer base includes major companies such as Netflix, Adobe, and Stripe.
– end
