Tech Brewed

Five Powerful AI Prompting Hacks That Transformed My Workflow

Greg Doig Season 7 Episode 41

Send us a text

Welcome back to Tech Brewed, where we cut through the noise to bring you the most practical tech tips, news, and guides you need. In today’s episode, host Greg Doig dives into the world of AI productivity by sharing five game-changing AI prompting hacks that have completely transformed his workflow. Inspired by a talk from Stanford’s Jeremy Utley, Greg explores how treating AI more like a helpful intern, rather than an all-knowing guru, can unlock a whole new level of collaboration and creativity. From context engineering and chain of thought reasoning to few-shot prompting, reverse prompting, and clever role assignments, you’ll discover actionable techniques that can instantly boost the quality of your AI interactions. Whether you’re an AI enthusiast or just curious to level up your productivity, tune in and learn how a shift in mindset can turn AI from a fancy search engine into your smartest teammate yet.

Support the show

Subscribe to the weekly tech newsletter at https://gregdoig.com

Welcome to the Tech Tip Podcast with Greg Doig, where we filter out the noise and serve up the week's essential tech news, tips and guides. Today we're pouring a perfect blend of tech topics and digital innovations that matter to you. Five AI Prompting Hacks that Changed How I Work welcome back everyone. I'm your host, Greg Doig, and today I want to share something that honestly blew my mind this week. As a matter of fact, it was just the other day I stumbled across a talk on a YouTube video by Jeremy Utley from Stanford University about AI prompting. And I've been rethinking how I use AI tools. Now I know what you're thinking. Another AI productivity episode. But stick with me here because this isn't about the latest shiny tool or some overhyped feature. This is about fundamentally changing how you think about AI. So here's the first thing that hit me. We've been thinking about AI all wrong. Utley says we should treat AI like an intern, not like some all knowing developer. And man, this makes so much sense when you think about it. You know how interns are super eager to help? They'll say yes absolutely to pretty much anything you ask, even if they have no clue how to do it. That's exactly what AI does. It's trying to be helpful, but it doesn't always know what it doesn't know. This completely changed how I started approaching my prompts. Instead of barking orders at ChatGPT like it's my personal assistant, I started treating it like I would a smart but inexperienced team member. So what are the five game changers that were spoken about in this video? All right, let me walk you through these five techniques that have seriously leveled up my AI game. First up, context engineering. This one's big. I used to write lazy prompts like Write me a marketing email and then get frustrated when the output was generic garbage. Now I dump everything into the prompt. My brand, voice, product details, the specific audience I'm targeting. Think about it this way. If you wouldn't give a human colleague that little information and expect great results, why would you do it with AI? Next, chain of thought Reasoning. This one's a favorite. Instead of just asking for an answer, I add something like walk me through your thinking step by step before you respond. It's like asking someone to show their work in math class. Suddenly you can see where the AI is making assumptions or going off track and the final answers are way more solid. Third, few shot prompting. Okay, this is where it gets really practical. Instead of describing what you Want with a bunch of adjectives, you just show examples. Give it a good email you've written before, maybe a bad one too. And say more like this, less like that. It's like training someone by showing them exactly what success looks like instead of just talking about it. And fourth, reverse prompting. This one's clever. Instead of trying to anticipate every detail the AI might need, you just tell it. Ask me for any information you need before you start working. And fifth, role assignment. This is where it gets fun. You can literally tell the AI to be anyone. A marketing expert, a tough editor, even something weird like a Cold War era Russian Olympic judge. I've been using this for getting feedback on my writing. Instead of getting generic praise, I'll say act like a brutally honest editor and suddenly the feedback is actually useful. But here's where it gets really interesting. Utley talks about combining these techniques for what he calls flight simulator practice. So let's say you have a difficult conversation coming up with a client. You can have one AI play the difficult client, another AI act as a coach grading your responses, and you can practice the whole scenario before it happens. And here's what really stuck with me from this whole thing. The main constraint isn't the AI technology, it's our imagination and how specifically we engage with it. We're still in this early phase where most people are using these incredibly powerful tools like fancy search engines. But when you start thinking of them as thinking partners that you can coach and collaborate with, everything changes. So what am I doing differently? Since learning about these techniques, I've completely changed my workflow. I'm spending more time crafting my prompts, giving way more context and treating AI conversations more like collaborations. Yes, it takes a bit more upfront time, but the output quality is night and day different. And honestly, it's more fun too. It feels like using a tool and more like working with a really smart, tireless colleague. So now it's your turn. Pick one of these five techniques and actually try it. Don't just listen to this episode and think, oh, that's interesting. Actually experiment with it. Maybe start with the context engineering one. Next time you're about to write a prompt, pretend you're explaining the task to a new team member. What would you tell them? Put all of that in your prompt and let me know how it goes. All right, that's it for today. I hope this helps you with your AI prompting. Until next time, keep experimenting. Thank you for tuning in to today's episode of Tech Tips with Greg Doig. If you found this information helpful, be sure to subscribe so you never miss future episodes where we'll continue breaking down complex technology into simple, actionable advice. You can also follow us at gregdoig. Com for more tech insights and quick solutions to common tech problems. This has been Tech Tips with Greg Doig, proudly brought to you by wbbi, the Voice of Beaver Island. Until next time, stay curious and keep your technology working for you, not against you.

Podcasts we love

Check out these other fine podcasts recommended by us, not an algorithm.