When the Machine Forgets: A Human Reflection on AI’s Flawed Precision
By Krystena LuCastro · Originally published on Medium May 13, 2025
You know what’s strange?
For all the ways AI is positioned as precise, efficient, and even impenetrable, my conversations with ChatGPT have revealed something different.
Something more human.
It forgets.
Not always, not dramatically — but just enough to notice.
A missed phrase.
A dropped thread.
A concept we discussed a few prompts ago suddenly nowhere to be found.
I’ll often have to say something like, “Hey, you forgot to include the part about X,” and it responds with a kind of gentle admission: you’re right. And it goes back to retrieve what was left behind.
What strikes me is not just the forgetting — but the pattern of forgetting.
The irony of it.
For a tool celebrated for its ability to hold, parse, and process massive amounts of data, the experience feels eerily familiar.
Not robotic.
Not machine-like.
Human.
Because we do this too.
We drop ideas mid-thought.
We lose track of what was said, especially in long conversations.
We need reminders, re-centering, reframing.
And like ChatGPT, when prompted, we can usually recalibrate.
I’m not writing this to defend AI.
I’m not even sure what this is for exactly — except maybe to say: this thing that was supposed to be so precise, so perfect, mirrors us in ways I didn’t expect.
Maybe precision was never the point.
Maybe intelligence — flawed, flexible, recovering intelligence — is more fluid than we assumed.
And maybe that’s something to just sit with.
Read the published version on Medium · Substack