Meta’s AI Race: Slow Coding, Fast Paychecks, And Vibe Fails
Meta’s big AI bet hit a speed bump—turns out their own systems, designed for billions, take 'too long' to deploy changes (hours vs. minutes) and kill the new trendy 'vibe coding.' To fix this, Meta’s Superintelligence Labs led by ex-GitHub CEO Nat Friedman adopted Vercel and GitHub in September 2025, shaving deployment times from 99 to 2 minutes. Fun fact: both Friedman and Meta’s AI chief Alexandr Wang are investors in Vercel, which recently raised $300M at a $9.3B valuation. Meanwhile, Zuckerberg was spotted rocking Meta x Ray-Bans, probably calibrating his hype glasses for the Silicon Valley AI showdown.
Share the Story
(1 of 3)Meta pays nine-figure salaries for top AI talent but still can’t vibe code on its own platforms—buying speed with Vercel and GitHub like a coder who can’t boil water without a blender.
Meta’s internal systems are so slow ‘vibe coding’ is impossible, so executives invested millions in another company that basically does Meta’s job faster and better. Efficiency via outsourcing the obvious!
When your internal AI tools are too slow for today’s culture of cool coder vibes, just spend hundreds of millions, hire ex-GitHub heads, and pretend to innovate by borrowing someone else’s tech.
Source: Businessinsider | Published: 10/3/2025 | Author: Pranav Dixit
More Articles in Technology
Man Applies to 200 Jobs with AI's Help, Still Does Own Eyerolling
KPMG’s Consulting Business Dies Dramatically While Staff Take a Hike
Motorola Razr Flip Phones Finally Told They’re Secure Enough for 2027
AI Boss Warns Robots Might Steal Liberty, But Only If Companies Don’t First
Grand Theft Auto VI Leaks So Hard They Might Just Skip Launch Party
French Techie Breaks AI Watermarks Faster Than It Takes To Say Bonjour
Mystery AI Appears, Impresses Silicon Valley, Nobody Asked Who Made It
