---
slug: up-to-what-extent-do-you-wish-to-follow-every-rabbit-hole
title: Up to what extent do you wish to follow every rabbit hole?
category: Notes
status: Published
summary: The problem starts when your cortisol level becomes directly proportional to your craving for fact-checking.
publishedAt: 2026-07-12
---

# Up to what extent do you wish to follow every rabbit hole?

One thing I truly hate when AI is working for me, whether it is polishing an email, generating RCAs for tech-related cases, or even helping me craft a response in MS Teams (kasi minsan medyo sabaw iyong kausap mo hehe), is the em-dash.

Yes, the em-dash.

Whenever I read emails or even casual responses and see too many of them, I immediately know. Yuck! (Joke hahaha.)

Of course, we will never really know if it was intentional. As far as I know, em-dashes existed long before the AI agents we use at work today. So yes, benefit of the doubt. (Problema baka puro doubt dios ko po!)

This time, I found myself working on a personal project: tracing facts and figuring out where things came from. But I also realized that I needed to set a border on when to stop.

The keyword is *border*.

Not when you feel tired of using your brain.

Back in 4th year high school (for those from the K12 generation in PH, that would be around 10th grader), we had a short story in Filipino class titled *Kesa and Morito*, translated by Lualhati Bautista.

What stayed with me all these years was not the author's name. It was the story itself: love, murder, and human behavior. As we all know, once love gets involved, humans start making very interesting decisions.

The story somehow ended up in my mental manifest.json because I easily get hooked on narratives that revolve around choices and consequences.

**So what bothers me now?**

During the age of AI, I became curious about where the story came from.

I started digging and eventually found Ryūnosuke Akutagawa, the Japanese author behind the version that influenced the translation I had read. But as soon as I found Akutagawa, I discovered something else: some of his works were themselves inspired by older repositories.

Then another rabbit hole appeared.

Was Akutagawa writing a fake story?

Of course not.

My question simply became: where did Akutagawa get the idea?

Then I caught myself.

This is exactly how many of us operate in the AI era.

We tell ourselves that we should never stop while doubt still exists. We keep tracing sources, validating claims, and opening new tabs until our browser starts questioning our life choices.

There is no doubt that Akutagawa wrote the version I was looking for. There is no argument there.

The problem starts when your cortisol level becomes directly proportional to your craving for fact-checking.

That is when you need to establish a boundary.

I told myself:

`This is what I understand at the moment`

The version I know is the Bautista translation. The upstream source I was able to verify is Akutagawa. Beyond that, things become more complicated because I do not understand Nihongo and those kanji write-ups. Tracing Bautista back to Akutagawa was already a challenge. Tracing Akutagawa further back opens another set of questions.

Sooner or later, I will probably end up tracing another repo, then another repo behind that repo, until I somehow find myself studying Japanese history instead of finishing this blog entry. At that point, I would not be surprised if I accidentally reach the Heian period and start investigating literary ancestors of literary ancestors while completely forgetting why I opened the browser in the first place.

That was when I decided to park the topic not because the rabbit hole was not interesting, but because the rabbit hole was working exactly as intended.

And if language becomes the barrier, well... the only Japanese phrase I know is *Arigatō*. There is another one, but mentioning it would probably derail this blog site faster than the Heian rabbit hole, so let us move on (yeah!).

It does not mean I am stopping.

It simply means I am parking the topic.

And perhaps that is the message of this journal.

Whenever we use AI, we should not stop the moment an output is delivered as if it were a Shopee, Lazada, or SHEIN parcel (this is our version of Amazon parcels).

Open it.

Inspect it.

Validate it.

Question it.

But at the same time, understand that every source has another source behind it, and every answer can lead to another rabbit hole.

Bautista is not fake.

Akutagawa is not fake.

The challenge is knowing how far you are willing to go before the pursuit of certainty starts consuming more energy than the question itself deserves.

For now, I am parking *Kesa and Morito*.

Maybe I will revisit it someday, maybe not.

But if there is one thing I learned from this exercise, it is that reading comprehension in the age of AI is no longer just about understanding what a text says.

It is also about understanding where it came from.

And while we are at it, please tune your AI agents to get rid of excessive em-dashes.

Be human sometimes (Kahit minsan lang bes).