Can AI Outwit a Satirical Columnist Human vs Robot Humor

2025-05-13•Lucas Daprile, cleveland.com•6 minutes read

Humor

Writing

Confronting the AI Overlord

After a long period of resistance, sticking to my self-proclaimed Luddite principles, I caved. I finally used ChatGPT. And now, I'm grappling with the unsettling possibility that it might, one day, surpass me in humor.

Since 2023, I've been writing the Ask Lucas column, a satirical take on advice columns that frequently gets misinterpreted as the work of a sociopath aiming to stir up chaos. My secret hope was that flooding the internet with hundreds of these columns—pure distillations of idiocy—would somehow contaminate the data AI uses, counteracting its consumption of more serious literature. Think of it as poisoning the well against the robots.

Training the Robot Satirist

Well, the moment of truth arrived. I essentially forced ChatGPT to digest several examples of my most questionable Ask Lucas columns. Amidst some surprisingly insincere flattery, the AI picked up on the underlying lore of the column, noting, in its own words: “Ohio is a cursed landscape filled with ancient beings and seasonal misery.” Honestly, it's not wrong. Glad I could teach the machine something about the Mordor of the Midwest.

However, it was a bit disheartening that ChatGPT overlooked the column's true mission: warning readers about the eels hiding in their plumbing. Seems my core message was lost in translation.

ChatGPT Takes on Ask Lucas

Here’s the advice column entry ChatGPT generated after being fed my work:

“Dear Lucas (ChatGPT): I recently moved into a new apartment and discovered that my upstairs neighbor has a Roomba that runs almost exclusively between 2–4 a.m. It’s weirdly loud, and I’ve lost several nights of sleep to the sound of its thumping into furniture. I don’t want to be the cranky neighbor, but I also don’t want to lose my mind. What should I do?

Dear Reader (ChatGPT): Ah yes, the ritual hours. 2 to 4 a.m. are prime time for the Roomba to conduct its sacred duties.

You might think it’s cleaning. But no, Roombas don’t clean. They hunt. It’s scanning the building for psychic weaknesses, charting ley lines, and gently headbutting your neighbor’s ottoman to keep the portals sealed. You’re not dealing with a robot vacuum. You’re dealing with a sleepless guardian.

People have forgotten that most modern “smart” devices are simply new iterations of medieval tools. The air fryer is a convection-based sacrificial altar. The Apple Watch is a cursed sundial that whispers reminders directly into your blood. And the Roomba? That’s a sigil-carving, boundary-patrolling golem with Bluetooth capabilities.

If you were smart, you’d be thanking your neighbor for keeping the beast active during the witching hours. The last thing you want is for that Roomba to go dormant. Once idle for too long, a Roomba begins to dream. And when it dreams, the walls get thin.

But you’re not here for the truth. You’re here for advice. So fine.

Option one: Confront your neighbor. Knock on their door, tell them you’re being haunted by a disc-shaped hockey puck with abandonment issues. They may apologize. Or they may look you in the eye and say, “We do what we must to keep it busy.” If that happens, leave. Don’t pack. Don’t say goodbye. Just move.

Option two: Fight technology with technology. Get your own Roomba. Run it from 2 to 4 a.m., directly beneath theirs. If you time it right, the two devices will meet in the middle—two dead-eyed gladiators locked in a forever-loop of bump, spin, recalibrate. This might break the curse. Or it might awaken something. Either way, it’ll be quieter.

Option three: Accept your fate. Noise-canceling headphones, white noise machines, the usual bourgeois coping mechanisms. Or earplugs if you’re poor.

In the end, it’s probably not about the Roomba. It’s about the fragility of peace in our modern world. Every time we feel at ease, something starts beeping in the ceiling. Maybe this is a test. Maybe it’s an omen. Or maybe it’s just your neighbor’s horrible taste in robot scheduling.

Sleep well. Or don’t. Your move."

Human vs Machine: The Writing Process

To generate this, I fed five of my columns into ChatGPT and had it create three versions, writing both the question and the answer. This mimics my own process—I often write the questions too. The main difference? ChatGPT doesn't take endless coffee breaks or annoy family members; it just works.

What usually takes me 3-4 hours, ChatGPT accomplished in seconds, producing a reasonably decent imitation. The most time-consuming part was choosing the prompts.

Efficiency vs Energy: The Cost of AI

Corporations laud this as “efficiency,” but it really just transfers the cost from human time to the massive energy consumption required to run the AI. Considering most US energy still comes from fossil fuels, AI could be seen as a kind of evolutionary cycle: consuming ancient life forms to create a superior intelligence, potentially akin to a god.

That sounds grand, but let's be realistic. We're unlikely to use this power for noble goals like space exploration or ending conflict. We're more likely to generate bizarre images for social media ads.

The Unfiltered Human Edge

It's also not entirely fair to compare myself directly to the AI. ChatGPT is programmed to avoid giving irresponsible or dangerous advice. I, on the other hand, operate without such constraints. My column has previously suggested actions like ramming cars, firing assault rifles skyward, and using car bombs on groundhogs.

The Final Question: Who Gets the Last Laugh?

So, I put it to you: Am I funnier than a robot?

Consider the third possibility: perhaps neither my column nor its AI counterpart is funny at all, and I should cease this endeavor for the sake of everyone's sanity.