Hello,
Once upon a time, Jensen Huang said that IT departments will become AI agents' HR managers.
Here we are.
In 2025 and before, agentic AI was more about prompts and tools.
In 2026, it's about orchestration.
It's now been a few months since OpenClaw went out - and we can say it wasn't just a hype. The open source Personal AI Assistant now has more than 325K GitHub stars, making it one of the most popular of all time.
What concretely made OpenClaw more popular than other agentic tools?
It's all about orchestration:
- smooth user experience (chat interface)
- memory management
- subagents
- connection with third party tools & APIs
- skills and features
It's not about an AI model. It's about how AI interacts with its environment and iterates with itself.
And this is far from being simple.
We can see OpenClaw competitors shipping every day to better orchestrate their agents. Claude Code just shipped channels while Codex just shipped subagents.
Software developers build their own agentic workflows and there are infinite possibilities to explore. IDEs are progressively transforming into AI agents orchestrators, like the Codex app or Conductor, a product that allows to switch between Codex and Claude Code agents in isolated workspaces.
This orchestration approach applies to every field, not only coding. Choose any industry and you can build an agentic software. It'll require some expertise about this field to be efficient.
For example, MyTrainer mobile app manages your training schedule from a chat interface based on your goals, preferences and even health metrics but it wouldn't be efficient if not built with fitness coaches.
There was a time where we thought ChatGPT would become the everything app, replacing all existing software. Then OpenClaw came in, then OpenAI hired Peter Steinberger..
It seems like AI enables people to build incredible things, incredibly fast. It didn't remove the importance of the right approach and original ideas.