Possible: Amjad Masad on vibe coding, AI agents, and the end of boilerplate

The most successful programming interface of the next decade won't look like code at all. It will look like having a conversation about what you want to build, then watching an AI agent construct it in real time while you provide feedback and direction. This isn't speculation, it's already happening, and Amjad Masad has built the platform proving it works.

Masad, the CEO of Replit, has spent the last few years watching non-programmers build sophisticated applications through what his team calls "vibe coding." The results challenge everything the software industry believes about who can and should write code. "There's a woman that had a very rare eye disease and required her to do certain exercises, and she built an app that helps her do these exercises with her eyes every day. There's a woman in Korea and her kid has a very rare disease, and she built an app to manage that disease on a day to day basis," Masad explains. These aren't programmers. They're people with problems who now have the tools to solve them directly.

The technical breakthrough enabling this shift comes from treating AI coding like a video game rather than a deterministic process. When Masad's team designed Replit Agent, they borrowed mental models from roguelike games like Hades, where each playthrough generates different experiences. "AI is stochastic. Even the language we use inside the company, an agent run is like a game run," Masad notes. This framing unlocked a crucial insight: instead of trying to make AI coding perfectly predictable, embrace the randomness and use it as a feature.

The practical implementation involves running multiple parallel attempts at solving the same coding problem, then selecting the best result. "You can take a file system, you can fork it a 100 times, run the same prompt with different parameters on these different forks, and then pick the best one," Masad explains. This approach works because Replit's infrastructure makes creating and discarding these experimental code environments essentially free. Most traditional development tools can't do this, they're too heavy, too slow, and too expensive to spin up dozens of parallel coding attempts.

The implications extend far beyond hobbyist projects. Masad identifies a specific business function ripe for disruption: revenue operations teams. "Those people that are managing a lot of the data flow within go to market teams and building the tools and helping salespeople do their jobs...tend to use a lot of SaaS software, but they don't have anything to connect all these things together to create applications and make their salespeople more successful." These teams currently stitch together Salesforce, HubSpot, Slack, and dozens of other tools through manual processes and basic integrations. With AI agents, they could build custom applications that unify these systems exactly how their business needs them to work.

This represents a fundamental shift in who controls software creation. For decades, businesses have been constrained by what existing SaaS products offer, plus whatever their engineering teams have bandwidth to build. The new model lets domain experts, the people who actually understand the business problems, create their own solutions without learning traditional programming.

Masad argues this democratization is inevitable and necessary. "My advice would apply to 99% of people. I don't think most people who wanna make things with software and computers don't need to learn how to code. It's gonna be a huge cost on them. It's gonna take a lot of time. But instead, just jump into making things." This isn't about replacing professional developers. It's about expanding the population of people who can create software from thousands to millions.

The transition won't happen overnight, but the early signals are clear. Non-technical users are already building functional applications through conversation interfaces. These tools are becoming more capable every month. The economics favor this approach, it's faster and cheaper than traditional development for many use cases.

Watch for this shift in three places: revenue operations teams building custom integrations, healthcare workers creating specialized tracking tools, and small business owners automating their unique workflows. When these groups stop asking their IT departments for simple applications and start building them directly, the transformation will be undeniable. The question isn't whether this future arrives, but how quickly your organization adapts to a world where anyone can code.