# Liwei 2min: Do Heroes Make History, or Does History Make Heroes?
I once tried to organize my granduncle's handwritten poetry manuscripts. Several times, I started and stopped.
To call it "organizing" isn't quite right.
It was more like resurrecting.
Because the materials had always been there.
Not discovered yesterday.
Not neglected.
And certainly not because no one knew how to organize them.
Quite the opposite.
The more precious something is, the more likely it becomes an abandoned project.
Because it's too much trouble.
Several hundred classical poems.
Handwritten drafts, typed copies, fragments — all mixed together.
OCR.
Collation.
Cataloging.
Annotation.
Digitization.
Just thinking about it gives you a headache.
So over a decade passed.
They just sat there. A permanent work-in-progress.
Like so many family genealogies.
Old photographs.
Memoirs.
Graduation yearbooks.
You know they're precious.
But they never make it to the top of the priority list.
Then one day.
Agents arrived.
And I suddenly realized: the projects hadn't gotten simpler.
The cost structure had changed.
What used to take one person half a year, or a full year.
Now takes a few evenings.
Projects that had been left in the cold now have the conditions to come back to life.
For this, we have Peter to thank. His OpenClaw opened the era of AI agents.
Many say Peter changed everything — that he made history.
I think this question resembles that old one from history class:
Do heroes make history.
Or does history make heroes.
Without Peter, would we still have today's agent wave?
Maybe a bit later.
But probably not absent.
Because what truly matured wasn't any single person.
It was the entire era.
The models matured. Especially Chain-of-Thought reasoning and reinforcement techniques reaching maturity and wide adoption — the essential foundation for agents handling long-running tasks.
The tools matured. Especially coding capabilities, and the ecosystem of skills and tool invocation.
Costs came down.
Context windows grew longer. Which meant expanded working memory.
Memory management emerged. Forgetting and dreaming mechanisms kept long workflows from dropping the ball.
And suddenly, a huge number of things that couldn't be done before — that weren't worth attempting — became doable.
So I increasingly feel:
Peter didn't invent the continent.
The continent was already there.
When the compass, the charts, and the ships are all ready, someone will set sail.
If not Peter.
Then John.
Or some Zhang San or Li Si.
History is always like this.
History calls forth heroes. Heroes push history forward.
What's worth savoring is this: this time, the hero wasn't a top AI scientist from a major-model frontier lab. He was a grizzled systems engineer.
by Tuya
