The user created a prompt hub to save, manage, and revisit prompts. The user was tired of losing their best prompts.
**I tested 100+ AI prompt hacks, only these 10 actually work.** I’ve spent the last few months testing every “viral” AI prompt trick I could find. TikTok hacks, Reddit posts, YouTube gurus—you name it. Most are fluff. But a handful actually deliver **consistently powerful results** across work, learning, and personal projects. Here are the 10 that made me stop and think, *“Wow, I was using AI wrong this whole time.”* **1. “Explain it like I’m 12, then like I’m an investor”** The comprehension sandwich. Example: *“Explain a content marketing funnel like I’m 12, then like I’m a CMO deciding how to allocate budget.”* You get both clarity and strategy-level depth in one shot. **2. “Act as a \[specific expert\] with years of experience”** Generic = generic. Specific = magic. Example: *“Act as a career coach who has helped 500+ people transition into tech from non-tech jobs.”* Way sharper advice than just “act as a coach.” **3. “Give me 3 options: creative, converting, and action oriented”** This is the decision spectrum trick. Example: *“I need to redesign my landing page. Give me 3 options:* creative, converting, and action oriented\*.”\* Now you see a range, not just one random suggestion. **4. “Before answering, ask me 3 clarifying questions”** For vague problems, this saves hours. Example: *“Help me pick a side hustle. But before answering, ask me 3 clarifying questions.”* AI narrows it down like a consultant instead of guessing. **5. “Think step-by-step and show your reasoning”** AI gets smarter when forced to walk through logic instead of spitting answers. Example: *“I’m building a Facebook ad campaign. Think step-by-step and show your reasoning for choosing targeting, copy, and creative.”* Instead of vague tips, you get a full breakdown of the thought process. **6. “What would \[specific thinker\] say about this?”** Perspective-shifting unlocks gold. Example: *“What would Steve Jobs say about this product idea?”* AI channels real-world patterns from training data, not vague platitudes. **7. “Challenge my assumptions and suggest alternatives”** Forces AI to push back instead of just agreeing. Example: *“Here’s my business model: \[insert details\]. Challenge my assumptions and suggest alternatives.”* Suddenly, AI is more like a critical co-founder. **8. “Summarize, then expand”** Compression + expansion gives sharper insights. Example: *“Summarize this 20-page report in 5 bullet points. Then expand each into actionable strategies for a startup.”* The two-step flow gives clarity first, then depth. **9. “Brainstorm 10 ideas and rank them by effort vs reward”** Not just brainstorming—structured prioritization. Example: *“Brainstorm 10 newsletter growth ideas and rank them by effort vs reward.”* Instead of chaos, you get a decision-making matrix. **10. “Combine perspectives”** Ask AI to blend roles = creativity unlocked. Example: *“Explain AI safety like a professor, a comedian, and a startup founder—all in one answer.”* The mashup sparks ideas no single perspective would. ⚡ **Pro insight**: Don’t stack all 10 at once. Pick 2–3 that fit your situation. AI thrives on context, not overload. I was tired of losing my best prompts, so I built a **Prompt Hub** to save, manage, and revisit them anytime. It’s free, and you can explore other curated prompts too. 👉 Try it here: [AISuperHub Prompt Hub](https://aisuperhub.io/prompt-hub) Which of these 10 have you tested—and what’s the most unusual role you’ve ever asked AI to play that actually worked?