I Made Money Just by Typing Funny Sentences for AI Testing 🤖💸

It sounds almost unbelievable: getting paid to sit behind a keyboard and type out funny, random sentences like “The chicken wore sunglasses while applying for a job as a lawyer.” Yet, that’s exactly what I stumbled upon—a quirky but very real way to earn extra cash online. The world of artificial intelligence has exploded, and companies need huge amounts of data to make these systems smarter, wittier, and more “human.” And guess what? They’re willing to pay people like me to generate some of the weirdest, funniest sentences imaginable.

In this article, I’ll share how I discovered this unusual gig, what it felt like to literally get rewarded for nonsense, why AI needs our humor, and how this small side hustle reveals bigger truths about the future of technology and creativity. So, buckle up—because this is the story of how I earned money just by typing funny sentences for AI testing.

The Day I Discovered the Weirdest Job Offer Ever

I wasn’t actively looking for a job that would pay me to be silly online. At first, I thought it was a joke—like one of those spammy ads that promise you can “earn thousands by doing nothing.” But curiosity got the better of me, and after digging deeper, I realized this was legit.

AI research companies and startups are in desperate need of something called training data. To build a chatbot that can respond with humor, sarcasm, or creativity, it must be exposed to thousands—sometimes millions—of quirky examples. That’s where people like me come in. We are the human suppliers of weirdness.

The task description was simple:

Write short, funny sentences that don’t necessarily make sense.
Keep them original, no copy-pasting.
Mix randomness with humor.

And the pay? Modest, but surprising—anywhere from $5 to $20 per task batch, depending on length and quality.

For someone like me who spends hours a day thinking of random jokes to annoy friends, this was heaven.

Why AI Needs Our Humor

It might sound strange that powerful AI systems need something as trivial as “dad jokes” or absurd one-liners. But humor is one of the hardest things for machines to understand.

Think about it: humor is built on context, culture, timing, and even shared human experience. A computer doesn’t inherently “get” why it’s funny to say:

“My refrigerator just texted me, ‘Stop opening the door, I’m cold!’”

But if we feed enough creative examples into its training process, it begins to learn the structure of a joke, the playfulness of language, and even cultural quirks that make people laugh.

I realized that my silly contributions weren’t just a waste of time—they were helping AI become more human-like. In a way, I was teaching robots how to laugh.

The First Batch of Funny Sentences I Typed

I remember my very first batch vividly. I was nervous, thinking: “What if my sense of humor isn’t good enough? What if the AI thinks I’m boring?”

Here are a few of the gems I submitted:

“The giraffe ordered a pizza and asked for extra pineapples, but only on Tuesdays.”
“My cat is writing her memoir, but she refuses to include the chapter about the vacuum cleaner.”
“I challenged my toaster to a dance battle, and honestly, it’s winning.”

After typing a couple dozen sentences like this, I submitted my work. A few days later, the payment landed in my account. The thrill wasn’t in the money itself, but in the bizarre joy of realizing: I just got paid for being ridiculous.

The Psychology Behind Typing Funny Sentences

The more I worked on this gig, the more I noticed something fascinating: writing nonsense sentences actually boosted my mood.

There’s research showing that humor stimulates dopamine, the brain’s “feel-good” chemical. By forcing myself to create absurd scenarios, I was essentially tricking my brain into laughing. And because I knew I was contributing to AI research, there was also a sense of purpose.

It wasn’t just “haha funny.” It was “haha funny—and I’m making the future of AI slightly sillier.”

The Surprising Challenges of Being Paid to Be Funny

At first, I thought it would be effortless. Who doesn’t like being goofy? But after a few hundred sentences, I realized humor has layers.

The hardest part was originality. I couldn’t just recycle the same structure like: “The dog did X while the cat did Y.” I had to keep it fresh, unexpected, and creative.

Some days, the ideas flowed like a comedy fountain. Other days, I stared at the screen for hours, trying to invent new scenarios: “What if… a penguin started a podcast about cooking?”

I learned that being paid to be funny turns humor into a kind of art form. It’s not about telling jokes; it’s about bending reality into something whimsical and surprising.

How Much Money Can You Really Make?

Let’s be real: this isn’t a full-time career. At best, I was making around $50–$100 a week, depending on how much time I spent and how many tasks were available.

But here’s the beauty of it—it was flexible, fun, and required no prior skills other than a quirky imagination. Compared to scrolling endlessly on social media, this was a better use of my free time.

Also, over time, I discovered multiple platforms running similar projects. Some were focused on humor, others on writing casual dialogue, and others even on testing sarcasm detection.

The more gigs I stacked, the more I realized: the future of micro-work is weirdly creative.

The Bigger Picture: Creativity in the Age of AI

What fascinated me most about this journey wasn’t just the money. It was the bigger question: What does it mean that machines are learning humor from humans?

Humor is deeply tied to culture. A joke that’s hilarious in one country may fall flat in another. By hiring diverse groups of people to write funny sentences, companies are essentially capturing a wide spectrum of human imagination.

I found myself part of something bigger—a global team of humorists training tomorrow’s AI to not only respond accurately but also with personality.

The irony hit me: by typing nonsense, I was shaping the future of intelligence.

When Funny Meets Philosophical

There were moments when I wondered: if AI learns humor from us, could it one day create jokes better than humans? What if an AI comedian sells out arenas?

The thought was both exciting and unsettling. On one hand, it shows the potential of technology to mirror human creativity. On the other, it raises questions about authenticity.

If a machine can crack a joke that makes you laugh until your stomach hurts, does it matter that it doesn’t “understand” the joke? Or is humor valuable precisely because it reflects the messy, human way we see the world?

Working on these tasks forced me to think about these deeper questions—while still laughing at the idea of “a banana running for president.”

A Day in the Life of an AI Humor Tester

Let me give you a snapshot of what a typical work session looked like:

Coffee ☕ (because nonsense requires caffeine).
Open the task portal and read the instructions: “Write 20 funny, original sentences.”
Start typing absurdity:

“The moon refused to shine tonight because it was feeling shy.”
“I wore socks on my hands to impress the pigeons at the park.”

Laugh at my own jokes (sometimes embarrassingly hard). Submit the batch and wait for approval.

It was almost meditative. I wasn’t just “working”—I was playing.

The Unexpected Benefits

Beyond the money and laughter, I noticed other benefits creeping into my life:

Better writing skills: Crafting punchy, witty sentences improved my overall creativity.
Stress relief: Humor writing became a form of therapy.
Community: Some platforms had forums where we shared our funniest creations, sparking laughter worldwide.

What started as a quirky side hustle turned into something much more rewarding than I ever expected.

The Future of Getting Paid to Be Funny

Will this kind of gig last forever? Maybe not. As AI gets better, it may eventually generate its own funny sentences. But here’s the catch: AI still needs human feedback, context, and cultural insight. Machines can imitate humor, but they don’t truly “live” it.

That’s why I believe there will always be a role for people in shaping how AI communicates with the world. And honestly, even if it fades, I’ll always treasure the time I got paid to write lines like:

“A llama applied for a Netflix account but couldn’t remember its password.”

✅ Sources

Binsted, K., & Ritchie, G. (1997). Computational humor: The JAPE system.
Mihalcea, R., & Strapparava, C. (2005). Making computers laugh: Investigations in automatic humor recognition.
West, S. M., Whittaker, M., & Crawford, K. (2019). Discriminating systems: AI and the future of work. AI Now Institute.
OpenAI Research Blog – Humor and language modeling experiments (2023).
Wired Magazine – The Future of AI and Humor (2022).

Written by the author, Fatima Al-Hajri 👩🏻‍💻