Similarity between your AI prompt and a 12-year-old

"Liam, clean your room," Sarah said, her voice echoing up the stairs.

Liam groaned. "Okay, Mom."

Hours later, Sarah peeked into his room. Clothes were shoved under the bed, his desk was a chaotic mountain of papers, and his bookshelf looked like a tornado had hit it. "Liam! What did you do?"

"I cleaned it," he mumbled, pointing to a small, relatively clear patch on his floor.

Sarah had simply said "clean," leaving Liam to interpret what that meant. For him, a small cleared area was "clean."

"No, you didn't! I meant everything. Make your bed, put your clothes away, organize your desk, and straighten your bookshelf," Sarah clarified, feeling a headache brewing.

She now listed specific tasks, breaking down the vague request into actionable steps.

Liam sighed. "Why? It's just going to get messy again."

He lacked the motivation to do a thorough job because he didn't understand the purpose.

"Because your grandparents are coming to visit next week, and I want them to see how responsible you're becoming," Sarah explained. "Plus, a clean room helps you focus on your homework."

Liam grudgingly started picking up his clothes. "Fine. But can I at least leave my robot models on my desk? I'm working on a design."

"Okay, but organize them neatly," Sarah relented.

An hour later, Sarah returned. The room was marginally better. Clothes were in the hamper, but not folded. The desk was organized, but the robot models were scattered haphazardly. The bookshelf was mostly straight, but with a few books sticking out at odd angles.

Sarah realized she needed to give more specific instructions even within the tasks.

"Liam, the clothes go in the drawers, folded. And the books should be flush with the shelf, not sticking out. And can you group the robot models by type?"

Liam rolled his eyes, but he started folding the clothes and arranging the books. "This is so much work," he complained.

"I know, but you're doing great," Sarah said, noticing he was genuinely trying. "Just a few more adjustments, and it'll be perfect."

Finally, after another half hour, the room was presentable. Liam, though tired and a little grumpy, looked at his organized space with a hint of pride.

"See? It wasn't so bad," Sarah said, smiling. "And now you have a clean room for your grandparents."

"Yeah, I guess," Liam admitted, collapsing onto his newly made bed. "But next time, can we just hire a robot to do this?"

Let's break this dialog down.

Here are some learnings that Sara gained from this interaction:

Misinterpretation: Sara failed to provide Liam the proper amount of specificity. When giving specific tasks, results improved.

Need for Context: Liam wasn't motivated to respond the correct way until he had the context of the request.

Need for Iteration: A 12-year-old often needs to iterate to grab the whole context and respond accordingly.

How does this apply to your AI prompt? The same keys will provide better responses from your prompt.

• Give specific tasks
• Provide Context
• Iteratively build a response giving tasking and context along the way.

These principles have changed my prompt engineering for good.

Next time you prompt, treat your prompt as a 12-year-old.

Jim Blizzard Principal "the office whisperer"