A billionaire CEO from Jordan just convinced Silicon Valley’s greybeards that talking to computers in plain English beats typing code—and now his $9 billion company is teaching Saudi schoolchildren to build software by describing what they want.
Story Snapshot
- Replit coined “vibe coding” in January 2025, formalizing a 2024 shift from typing code to conversing with AI in natural language
- The company’s $9 billion valuation rides on Agent 4 and Canvas tools that let non-coders create software, fashion designs, and marketing videos by describing ideas
- Legendary developers Ryan Dahl and Andrej Karpathy endorsed the approach, while Anthropic built Claude Cowork entirely via vibe coding in 10 days
- Enterprises face bottlenecks in security and operations despite rapid feature development, forcing competitors like Cursor and Cognition to build DevSecOps automation
- Replit targets global expansion in Saudi Arabia, India, and Japan, positioning vibe coding as an educational democratization tool for childhood learners
The Billion-Dollar Bet on Conversational Creation
Amjad Masad didn’t set out to eliminate programming as we know it. The Replit cofounder launched a code-sharing platform that evolved into something far stranger: a digital whiteboard where humans and AI collaborate by exchanging vibes instead of syntax. In 2024, Replit released its first tool enabling users to describe software in conversational English rather than JavaScript or Python. By January 2025, Masad gave the phenomenon a name—vibe coding—and Wall Street valued his company at $9 billion. The premise sounds absurd until you watch a marketing manager generate a promotional video or a fashion designer sketch garment variations without touching a line of code.
When the Old Guard Stops Laughing
Skeptics expected seasoned developers to mock vibe coding as a gimmick for amateurs. Instead, Node.js creator Ryan Dahl and AI researcher Andrej Karpathy publicly embraced spec-driven, vibe-first workflows. Their endorsement signaled credibility: if the architects of foundational technologies trust natural language prompts over keystrokes, the shift isn’t trivial. Anthropic underscored the point by shipping Claude Cowork—a collaboration feature—entirely through vibe coding in 10 days. That timeline would shame traditional development cycles, and it validated Masad’s claim that vibe coding accelerates innovation beyond incremental improvements. The greybeards aren’t laughing; they’re experimenting.
Beyond Code: Fashion, Art, and Digital Scans
Replit’s Agent 4 and Canvas tools extend vibe coding past software into creative domains. Users right-click objects on a digital canvas to request variations—a dress in different fabrics, a logo in alternate color schemes, or a slide deck reimagined for executive audiences. Masad describes uploading a photograph and receiving a digital scan suitable for 3D rendering or product mockups. This expansion transforms vibe coding from a developer shortcut into a universal creation interface. Marketers build video campaigns, artists iterate designs, and educators craft lesson materials by articulating intent rather than mastering specialized tools. The implications ripple across industries reliant on visual communication and rapid prototyping.
The Enterprise Paradox: Speed Meets Structural Inertia
Vibe coding mainstreamed quickly among startups and AI labs, but large enterprises stumble over their own complexity. While developers ship features faster using natural language prompts, downstream bottlenecks in security reviews, site reliability engineering, and product management choke velocity gains. Cursor and Cognition recognized the gap, ramping DevSecOps offerings and code review automation to integrate vibe coding into enterprise pipelines. Cognition recently upgraded its review tools, targeting companies that generate code rapidly but lack infrastructure to deploy it safely. The challenge isn’t technical—it’s organizational. Enterprises optimized for typed code workflows resist reimagining their entire software lifecycle, giving nimble competitors structural advantages despite smaller budgets.
Global Classrooms and Geopolitical Ambitions
Masad’s expansion strategy targets Saudi Arabia, India, and Japan, positioning Replit as an educational bridge for non-Western learners. Partnerships with Saudi leaders aim to integrate vibe coding into elementary curricula, teaching children to create software by describing problems rather than memorizing syntax. The initiative aligns with Masad’s immigrant narrative—a Jordan native who built a billion-dollar company and now champions accessible technology for underserved regions. His public commentary on Palestinian issues and Middle East partnerships inject geopolitical dimensions into what might otherwise appear purely commercial. Whether vibe coding democratizes creation or concentrates power in AI gatekeepers remains an open question, but Replit’s classroom push suggests Masad believes early adoption determines long-term influence.
The Vibe-First Future: Disruption or Distraction
Vibe coding’s rapid ascent from 2024 prototype to 2025 buzzword reflects AI’s broader trajectory: automating expertise previously reserved for specialists. Masad frames the shift as world-changing, enabling anyone to translate imagination into functional output. Conservative skepticism toward unproven paradigms makes sense—Silicon Valley overhypes novelties routinely. Yet the endorsements from Dahl and Karpathy, Anthropic’s 10-day production deploy, and enterprise scrambles to automate pipelines suggest substance beneath the hype. The real test isn’t whether vibe coding works—it demonstrably does—but whether it liberates creators or simply shuffles dependency from human experts to AI black boxes. Enterprises crawling toward adoption while startups sprint ahead reveals a familiar pattern: innovation favors the nimble, and legacy players pay catch-up premiums. Masad’s classroom gambit in Saudi Arabia and beyond bets that whoever teaches the next generation defines the rules.
Sources:
How Enterprises Can Adopt Vibe Coding
