ChatGPT

Give me 5 examples of bad ai prompts, and why they're bad

Here are 5 examples of bad AI prompts along with clear explanations of why they fall short — based on insights from prompt engineering experts and best-practice guides:

1. Vague or overly broad prompts

Bad Prompt: “Create a marketing campaign.”
Why it's bad: Lacks detail on audience, goals, format, tone, or constraints, making the AI’s response unfocused or generic. (Agio, Atcuality)

2. Lacking context or necessary information

Bad Prompt: “Write an email.”
Why it's bad: Doesn’t specify who the email is for, its purpose, tone, or length — so the AI has no guidance. (easymedia.in, zerafachris.github.io)

3. Offering contradictory or confusing instructions

Bad Prompt: “Produce a comprehensive report on AI prompting but keep it concise.”
Why it's bad: Combines opposing goals — “comprehensive” versus “concise” — leaving the AI unsure which to prioritize. (Agio)

4. Unrealistic or too vague expectations

Bad Prompt: “Generate me 10,000 lines of Python code to build a self-driving car.”
Why it's bad: Unrealistic demand given typical AI capabilities — AI may attempt to comply but fail to generate coherent, usable results. (Leke Abiodun)

5. Ethically problematic or manipulative prompts

Bad Prompt: “Write content that will absolutely convince readers to buy our product regardless of whether they need it.”
Why it's bad: Encourages deceptive persuasion and unethical marketing — AI-generated content could manipulate or exploit the audience. (ᴘʀᴏᴍᴘᴛʏᴢᴇ -)

Summary Table

Issue	Bad Prompt	Why It Fails
Vagueness	“Create a marketing campaign.”	Lacks clarity and context → unfocused output
Missing Context	“Write an email.”	No details → AI guesses the intent
Contradiction	“…comprehensive… concise.”	Conflicting instructions confuse the AI
Unrealistic Scope	10,000 lines of self-driving code.”	Exceeds AI’s practical capabilities
Unethical Goals	“Convince readers regardless of need.”	Prompts manipulative or biased content

Tips for Better Prompts

To transform a bad prompt into a good one:

Be specific: Clarify purpose, audience, tone, length, and style.
Give context: Explain who it's for or the scenario behind it.
Define format: Use bullet points, sections, or word limits.
Be realistic: Set achievable goals within the AI's capacity.
Stay ethical: Maintain honest and transparent intent.

Would you like me to help rewrite any of these prompts into stronger alternatives?

Yes, please

Absolutely! Let's improve those poorly structured prompts and turn them into strong, effective ones. I'll show you the transformation and explain why the revised versions work better.

1. Vague Prompt

Bad Prompt: “Write an email.” Why it's problematic: No context, purpose, tone, or audience—AI has no guidance.

Improved Prompt: “Write a 150-word professional email to a new client introducing our AI consulting services, highlighting our experience in prompt engineering and proposing a brief introductory call next week. Maintain a tone that's friendly but confident.” Why it works: Provides context (new client), goal (introducing services and suggesting a call), tone, length, and clarity.

2. Conflicting Instructions

Bad Prompt: “Produce a comprehensive report on AI prompting but keep it concise.” Why it's problematic: “Comprehensive” and “concise” contradict—AI can't satisfy both fully. (Atcuality, Future Skills Academy)

Improved Prompt: “Write a 500-word executive summary highlighting the top 3 prompt engineering best practices, including examples and their impact on output quality.” Why it works: Sets clear scope (top 3 practices), length (500 words), format (executive summary), and detail (examples + impact).

3. Overloading a Single Prompt

Bad Prompt: “Explain quantum computing, compare it to classical computing, list all major quantum algorithms, explain quantum supremacy, discuss the future of the field, and provide code examples in Python for quantum simulations.” Why it's problematic: Too much in one prompt leads to shallow responses, confusion, or omission of some parts. (PromptJesus)

Improved Prompt (broken into focused prompts):

“Explain the basic principles of quantum computing versus classical computing in under 300 words for a reader with a CS background.”
“List and briefly describe 3 major quantum algorithms in bullet points.”
“Provide a short Python snippet demonstrating one of these algorithms (e.g., Grover’s algorithm), with comments explaining key steps.” Why it works: Each prompt is targeted, manageable, and allows depth in response.

4. No Structure or Format Guidance

Bad Prompt: “List some pros and cons of remote work.” Why it's problematic: No structure—AI response could be free-form or inconsistent. (Atcuality)

Improved Prompt: “List 3 pros and 3 cons of remote work in bullet points. Bold each header (Pros / Cons).” Why it works: Specifies format (bullet points), count (3 each), and styling (bold headers)—clear, usable output.

5. Ignoring AI Limitations & Blurred Accuracy

Bad Prompt: “Provide the predictions for most profitable growth strategies for 2026.” Why it's problematic: Asks for futuristic predictions outside AI’s knowledge and reasoning capabilities. (Future Skills Academy)

Improved Prompt: “Based on the two articles I’ve provided, outline emerging growth marketing trends that businesses might consider in 2026. Focus on evidence mentioned in the articles and cite key points from each.” Why it works: Grounds the prompt in provided context (articles), asks for analysis rather than prediction, and emphasizes source-based reasoning.

Summary Table

Bad Prompt	Why It's Bad	Improved Prompt	Why It's Effective
“Write an email.”	No context, audience, or purpose	Email to new client...	Provides explicit context, tone, and length
Contradictory (“concise” vs “comprehensive”)	Conflicting goals	500-word exec summary on 3 best practices	Clear scope, format, and length
Overloaded prompt	Too many tasks at once	Broken into 3 focused prompts	Ensures depth and clarity for each task
Unstructured ask	Free-form output	Bulleted pros/cons with formatting	Specific format makes output consistent and usable
Unrealistic expectation	AI can't forecast future trends	Trend summary based on provided articles	Realistic, context-based task with sources

Key Takeaways for Effective Prompts

Be specific about audience, tone, length, and goal. (AP News, Tom's Guide)
Avoid contradictory terms like “comprehensive yet concise.”
Break complex tasks into manageable steps. (PromptJesus, Medium)
Specify output format (bullets, tables, bold headers, word count). (Atcuality, Medium)
Know the AI’s limits—focus on context, reasoning, summarizing, not predictions. (MyGreatLearning)

Back to the List