Why Your AI Assistant Sounds Generic (And How to Fix It)
The internet is abuzz with news and posts about how powerful AI assistants have become. Every day, articles declare these tools life-changing, world-changing, job-changing. You try it yourself: open an AI assistant, paste a draft of a report you are writing, ask the AI to improve it, and hit enter.”
You impatiently wait for the magic to happen. Instead, what you get back is generic fluff. Unhelpful. Unaligned with your work.
You wonder: is everyone exaggerating? Or am I missing something?
The missing ingredient, more often than not, is the prompt.
What Makes a Prompt Effective?
A prompt is how you communicate with the AI. It is how you tell the AI what to do. It can take the form of a question, instruction, or request. But here is the catch: many of us write prompts the way we would talk to a colleague who knows us well, someone with whom we have shared context and experiences.
Your AI assistant, however, lacks that history. It is not a mind reader.
Instead, think of your AI assistant as a highly trained consultant walking into your project cold. It is brilliant, but only if you brief it properly.
Models of yesteryears demanded precise wording and carefully framed prompts to produce meaningful results. Today’s newer generations are more capable and forgiving, but the core truth remains: a well-crafted prompt still delivers better outputs. As MIT Sloan’s article on their AI Hub, titled “Effective Prompts for AI: The Essentials”, notes, context and specificity are the keys to unlocking business-relevant results.
Prompt Design in Action
Let’s make this concrete with some business examples.
Example 1: CV Optimisation
Vague Prompt: “Improve my CV.”
What you’ll get: a generic rewrite with buzzwords that may not reflect your strengths.
Improved Prompt: “Act as a career coach and recruiter with expertise in tech leadership roles. Rewrite my CV to target senior product management positions in AI-focused companies. Highlight measurable achievements, leadership impact, and technical expertise. Structure it in a clear, results-oriented format with concise bullet points under each role. Keep it to two pages.”
✅ Why it works: The improved prompt assigns the AI a role (career coach and recruiter), specifies the target audience and goal (getting a senior product management role), and defines the desired structure and focus areas. This produces a tailored, actionable CV instead of a generic rewrite.
Example 2: Internal Communications
Vague Prompt: “Summarise this meeting.”
What you’ll get: a generic recap.
Improved Prompt: “Summarise this leadership meeting into a one-page executive briefing. Organise it into: (1) Key Strategic Decisions, (2) Risks Raised, and (3) Next Steps with owners. Keep the tone neutral, under 300 words, and use bullet points for clarity. This summary will be shared with senior leadership.”
✅ Why it works: The AI now knows the audience (executives), the format (one page, bullet points), and the priorities (risks and decisions).
The Jagged Frontier of AI Capabilities
Even with excellent prompts, AI has limitations. As Ethan Mollick notes in Centaurs and Cyborgs: On the Jagged Frontier of AI (2023), “On some tasks AI is immensely powerful, and on others it fails completely or subtly.” This unevenness of AI’s capabilities is what Mollick and his colleagues call the “Jagged Frontiers.” While AI can greatly boost productivity in certain tasks, it can still stumble on others, even when they seem straightforward to a human.
Better prompts will not eliminate these gaps. Use AI to accelerate drafting, brainstorming, and structuring work, but always apply your own judgment and expertise when evaluating the results.
Three Takeaways for Better Prompts
There is no magic formula for the perfect prompt. Crafting effective prompts is a learning process that improves with trial and error. These patterns can help you get better outputs as you experiment:
- Context is everything. Treat the AI like a skilled consultant who is new to your project. Provide relevant background, goals, and constraints so it can produce aligned results. Brief it clearly!
- Specificity drives clarity. Be precise in your ask. The more specific your request, the better and less generic the output. Specify the role the AI should take (e.g., ‘Act as a career coach’ or ‘Act as a project manager’) and the perspective or expertise it should use.
- Define success. Specify the desired length, tone, audience, and action. If you want the output in a particular format, say so! Bullet points, tables, sections, or another structure, be explicit about it.
You rarely get it right on the first try. Use the conversational nature of AI to refine your prompts and improve your results. AI is powerful only when prompted well. With clear instructions and an understanding of its jagged capabilities, you can produce outputs that are actionable, aligned, and professional.
The best way to improve with AI assistants is to use them. Experiment with different prompts, explore what they can do, and discover how they can help you.
Here is a challenge: Take this prompt ‘Write an email to a stakeholder about a new product release.’ How would you adapt it to make the output relevant and usable? Experiment, tweak, and share what works.
Happy prompting!