Peter Steinberger didn’t set out to build the most talked-about AI agent framework of 2026. He just wanted things that didn’t exist yet.
In the first episode of OpenAI’s Builders Unscripted podcast, Steinberger sat down with OpenAI’s Head of Developer Experience Romain Huet to talk about OpenClaw’s origin story, his approach to building with AI, and why he thinks people give up on AI-assisted development too quickly.
No Grand Plan
“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.”
He started by building a WhatsApp integration tool, then set it aside — assuming the AI labs would ship something similar soon. When they didn’t, he kept building.
The real inflection point came during a weekend trip to Marrakesh. With spotty internet, WhatsApp was the only thing that worked reliably. He found himself using his prototype constantly — finding restaurants, looking things up on his computer remotely, sending messages to friends.
“Where it really clicked was where I was at this weekend trip in Marrakesh, and I found myself using it way more because it was so convenient."
"Vibe Coding Is a Slur”
Steinberger’s sharpest take was on the term “vibe coding” — the idea that you can just describe what you want and AI will build it.
“I think vibe coding is a slur,” he said.
His point: the term implies a trivial, unskilled process. In reality, building effectively with AI is a skill that takes practice. People try it once, get disappointing results, and conclude it doesn’t work.
“They try AI, but they don’t understand that it’s a skill. You’re not going to be good at guitar on the first day.”
Over time, he developed intuition for prompt engineering — knowing how long a task should take, recognizing when something’s going wrong, and adapting his approach. That intuition doesn’t come from reading about it. It comes from doing it, repeatedly.
The Advice
Steinberger’s guidance to builders is straightforward:
“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.”
He’s also optimistic about what this means for high-agency people worried about AI replacing them:
“If your identity is: I want to create things. I want to solve problems. If you’re high agency, if you’re smart, you will be in more demand than ever.”
What This Means for the OpenClaw Community
A few takeaways for builders:
- Start with something you actually want. The best agent setups come from solving your own problems, not following tutorials.
- Expect a learning curve. Your first agent workflows will be rough. That’s normal.
- Iterate, don’t plan. Steinberger didn’t blueprint OpenClaw — he explored, tested, and refined.
- AI-assisted building is a skill. Treat it like any other skill: practice deliberately, reflect on what works, and improve over time.
The irony isn’t lost on anyone: the creator of the most sophisticated personal AI agent framework says the secret is to just play.
Sources: TechCrunch, OpenAI Builders Unscripted Podcast. Related: OpenAI hires OpenClaw creator, how to create your own custom skill.