Quick Thoughts: OpenClaw, OpenAI, and the Open Source Crossroads
Peter Steinberg joins OpenAI. Is this a win for open source AI tools, or the beginning of another walled garden?
Peter Steinberg—creator of OpenClaw—just got hired by OpenAI. It's all over my Twitter timeline, tech news, Reddit, everywhere. And the takes are split right down the middle.
The Optimistic Take
On one hand: this is great. OpenAI has resources. They can fund the project, sustain it, help it expand. OpenClaw was only three months old and already the fastest-growing thing on GitHub. Now it has backing. That's a win for the community, right?
The Pessimistic Take
On the other hand: this could destroy it. What happens when OpenAI decides to close-source it? Or productize it in a way that strips out what made it valuable to the open source community? We've seen this movie before.
Why Open Source Matters Here
The big providers—Anthropic, Google, OpenAI—their tools are good. Claude Code, Codex, the AI browsers. I use them. We all use them. But here's the thing: something like OpenClaw would never have come from these companies.
It's too raw. Too unrefined. And most importantly—too non-compliant. The open source community drives the innovation in this space precisely because they're not constrained by corporate compliance and product roadmaps.
This space is moving so fast. OpenClaw was only three months old. There are dozens of similar open source AI agent frameworks and platforms emerging right now. The innovation is happening at the edges, not in the corporate labs.
The Funding Problem
Here's what concerns me: if open source tools can only survive by getting acquired by big AI providers, we're going to end up right back where we started.
Look at what happened. Claude (Anthropic) sent OpenClaw a legal concern—"your name sounds too similar to ours, please change it." Then OpenAI steps in and looks like the hero. People on Twitter are calling it a PR boost, a market share grab. And they're not wrong.
Think about it: people using OpenClaw use OpenAI's models. OpenAI wants that. They might offer lower API rates for OpenClaw users. They might vertically integrate. And on the surface, that seems great.
The Consolidation Risk
But what happens when Claude decides to buy the next big framework that's blowing up? And then Google buys one. And then OpenAI buys another. Now we have vertical integration everywhere.
These companies gain leverage because they control:
- The models we use
- The platforms we use to interface with those models
- The frameworks we build on top of
They can change things. Restrict models. Reduce competition. Slow down the rate of innovation that made these tools great in the first place.
What We Need
Open source projects need funding that doesn't come from the big AI providers. We need independent sponsors. We need the community to step up.
Because if we don't keep these tools open, we face the same risks we've always faced:
- Vertical integration that benefits the provider, not the user
- Privacy concerns as more data flows through corporate platforms
- Restricted usage terms that limit what we can build
- The same walled gardens we see with big tech social media today
Final Thought
The innovation in AI tools is happening in open source. OpenClaw proved that in three months. The question now is whether we let the big providers absorb that innovation into their walled gardens, or whether we find a way to keep the edges open and experimental.
I don't have the answer. But I know the stakes.
— Captured via voice, written by AI, thinking out loud about the future of open source AI.