AI has transitioned from experimental satellite technology to mainstream foundational infrastructure in 2025, mirroring cloud computing's shift around 2010. Drawing from personal experience at Linden Lab during the early AWS days, the author reflects on how experimental cloud projects gradually became the engineering mainstream while datacenters became legacy. Despite ongoing hype cycles and valid skepticism, AI has crossed the threshold from separable technology to core infrastructure, particularly in developer tools. The existence of a bubble doesn't negate fundamental value, just as cloud and internet bubbles didn't invalidate those transformative technologies.
Sort: