---
title: "Stop Asking ChatGPT to "Make It Better""
description: "5 source-backed AI claims from Stop Asking ChatGPT to "Make It Better", with key statistics, context, and the original AI Adopters Club source."
url: "https://kbanc.com/claims-library/stop-asking-chatgpt-to-make-it-better"
source: "https://aiadopters.club/p/stop-asking-chatgpt-to-make-it-better"
date: "2025-05-04"
topics: ["strategy", "tools", "implementation"]
generated: "2026-08-31"
---

# Stop Asking ChatGPT to "Make It Better"

By Kamil Banc | May 4, 2025

## Claims

1. **Fed Study on AI Savings** (source summary): A 2025 St. Louis Fed study found generative AI users saved 5.4% of weekly work hours.
2. **Daily Users Save More** (source summary): Among daily AI users, 33.5% saved four or more hours weekly, compared to 11.5% of occasional users.
3. **Prompt Report Scale** (source summary): The Prompt Report analyzed 1,500 academic papers on AI prompting across 76 pages.
4. **The Vague Prompt Tax** (Kamil's interpretation): Vague prompts like 'make this better' waste time through iterations and miss business context.
5. **Prompts Clarify Thinking** (Kamil's interpretation): Filling in specific prompt details clarifies your own thinking about communication goals.

## Evidence

### Quote
> "The difference between average and exceptional AI use isn't the model. It's how specifically you instruct it." - Kamil Banc

### Key Statistics
- **5.4%**: Average share of weekly work hours saved by generative AI users, per a 2025 St. Louis Fed study
- **2.2 hours**: Weekly time savings on a 40-hour schedule equivalent to the 5.4% average savings
- **33.5% vs 11.5%**: Share of daily AI users saving four or more hours weekly versus occasional users, showing precision of use matters

## Context
The article grounds its argument in a 2025 St. Louis Fed study quantifying AI time savings and in The Prompt Report, a 76-page synthesis of 1,500 academic papers confirming that structured, specific prompts outperform vague instructions. Kamil Banc, a practitioner writing for his 'Adopter' newsletter audience, translates these findings into eight fill-in-the-blank prompt templates covering rewriting, editing, storytelling, social threads, explanations, comparisons, and outreach messages. Each template replaces a common vague request like 'make this better' with precise parameters for audience, tone, structure, and call to action. The methodology is deliberately practical: readers are told to pick one template, fill in the brackets with their own specifics, and observe the improvement in output quality. The practitioner application extends beyond AI outputs, as Banc argues the exercise of specifying details crystallizes the writer's own thinking about each communication goal.

## Source
- Original: [Stop Asking ChatGPT to "Make It Better"](https://aiadopters.club/p/stop-asking-chatgpt-to-make-it-better)
- Cite: kbanc.com/claims-library/stop-asking-chatgpt-to-make-it-better

## Primary Evidence
- [Prompt Report](https://x.com/learnprompting/status/1800931910404784380) (x.com; supports claim 3)