In this post the author argues that programming remains the key to escaping poverty, especially in an era where AI amplifies the power and reach of softwareâno longer needing warehouses or large teams, but instead enabling local AI to perform tasks such as converting images into 3âD models for VR worlds. He reflects on his own journey from a novice programmer with no milestones to now presenting a flowâbased system called âlulz,â a textâbased visual language that uses ffmpegâstyle filtergraphs in JavaScript array notation and borrows ideas from NodeâRED. The post highlights how AI can finish programs quickly, yet also shows the challenges of visual programmingâgrouping nodes, nesting, and managing cablesâwhich are hard mentally but fit well with AIâs grunt work. By describing a textâvisual approach where cable connections are explicit in code, he envisions an AI that listens to these flows, samples data at each link, and thus creates modular programs automaticallyâan early glimpse of the revolution he expects for AI coding.






















