OpenClaw Creator: Experiment and Play with AI

OpenClaw creator Peter Steinberger: Experiment, play, and iterate with AI

Peter Steinberger, the developer behind the viral AI agent OpenClaw who was later hired by OpenAI, has practical advice for people experimenting with AI technology — including AI agents. From his experience, the best way to build today is to explore, be playful, and accept that you won't be an expert overnight.

"I wish I could say that I had the unified plan in the beginning, but a lot of it was just exploration," Steinberger said. "I wanted things, and those things didn't exist, and…let's say, I prompted them into existence."

From WhatsApp idea to OpenClaw

Steinberger told OpenAI's Head of Developer Experience, Romain Huet, on the first episode of OpenAI's Builders Unscripted podcast that OpenClaw started without a detailed plan. He initially built a tool to integrate with WhatsApp but put it aside, assuming larger AI labs would soon offer similar capabilities.

He continued to experiment and focus on fun and inspiration. By November, when those capabilities still hadn't appeared from other teams, he created the initial OpenClaw prototype.

One turning point came on a trip to Marrakesh. With spotty internet, Steinberger relied on WhatsApp and his prototype, finding it convenient for tasks like finding restaurants, looking up information on his computer, and sending texts. That real-world usage demonstrated the practical value of integrating AI with familiar messaging platforms.

What he learned about modern AI models

As he played more with the technology, Steinberger noticed how strong modern AI models have become at problem-solving. They can propose workable solutions without being explicitly programmed for every step.

He also found his own workflow improved as he gained experience. He stresses that this improvement takes time and practice, and urges developers not to give up when early results fall short.

On learning AI-driven development

Steinberger criticized the idea that using AI to code should be effortless. He warned against treating "vibe coding" as a shortcut; the term can create unrealistic expectations. Instead, he likened learning to combine AI with coding to learning guitar — a new skill that improves with practice.

• Start playful: Approach prompts and prototypes like experiments.
• Build what you want: Solve problems you personally care about.
• Reflect and iterate: If a prompt takes longer than expected, analyze and adapt your approach.

"My… advice always is, approach it in a playful way. Build something that you always wanted to build. If you're at least a little bit of a builder, there has to be something on the back of your mind that you want to build. Like, just play."

Career perspective: creators remain in demand

In a moment when many worry about AI replacing jobs, Steinberger offered a counterpoint: people who create and solve problems will stay valuable. High-agency individuals with creative skills will likely be in even greater demand as AI becomes more widespread.

Why this matters to gaming news and developers

Although Steinberger's work focused on messaging and productivity, his insights apply to game developers and gaming news audiences. Experimenting with AI agents can unlock new design workflows, procedural content generation, and smarter in-game assistants. Game studios that adopt playful prototyping and iterate quickly can discover practical features that improve player experience.

About the reporter

Sarah Perez has worked at TechCrunch since August 2011, after over three years at ReadWriteWeb. Before reporting, she held I.T. roles across banking, retail, and software. Contact Sarah at [email protected] or via Signal at sarahperez.01 for verification of outreach.