For the past two weeks, I've been spending a lot of time working with my AI agent, OpenClaw.
On one hand, I'm constantly impressed by how powerful it is. Whether it's writing code, building websites, or handling automation tasks, it can accomplish things surprisingly quickly.
On the other hand, I've also discovered one obvious weakness: it forgets.
A feature we worked on yesterday may be completely absent from its memory today unless I remind it again. While this can be mitigated by storing information in files or building memory mechanisms, the problem still appears from time to time.
At first, I saw this as a flaw of AI.
Then I realized something.
Am I really any different?
Learning Japanese has taught me the same lesson. Words I studied yesterday may be forgotten today. Grammar I thought I had mastered often becomes blurry after a few days.
Much of learning seems to be an endless cycle of learning, forgetting, and relearning.
Perhaps humans and AI are not as different as we like to think.
AI relies on memory files to preserve context. Humans rely on notes, repetition, and practice to fight against forgetting.
What I find even more interesting is how easily we notice the flaws of others.
When AI forgets something, we notice immediately. When someone else makes a mistake, we are quick to point it out.
Yet when the same thing happens to ourselves, it is often much harder to see.
Maybe the real question is not why AI forgets, but whether we can extend the same understanding and patience to ourselves.
After all, both humans and AI spend their lives forgetting, learning, and growing.