Prompt Engineering Made Easy: Non-Technical Guide

YouTube tutorial (hosted by Igor) titled “Prompt Engineering Made Easy: Non-Technical Guide” that reframes prompt engineering as a translator-style, non-technical craft. The video teaches three unexpected but practical skills to improve prompts—philosophy (using frameworks like ethos/pathos/logos and the Socratic method), storytelling (narrative-driven prompts inspired by Steve Jobs’ approach), and creativity (role-play/imagery prompts)—and uses before/after prompt examples to show how each skill yields more specific, useful outputs. The transcript also includes a sponsor segment (Brilliant.org) and a plug for the speaker’s prompt engineering course.
Questions this video answers
July 11, 2023
Philosophize This
Building a Story Brand
One month free trial; first 200 users get 20% off the annual premium subscription.
Chapters
There's a lot of hype these days around AI and the new jobs that it's creating. And yeah, as you might have already guessed, I'm talking about prompt engineering here, which is the art of talking to AI models. Now, when you hear that term, you might be thinking engineer.
That sounds very technical and I'm not.
A techie, but hold up. I'm here to tell you that word has been picked incorrectly. Because in the context of AI, engineer isn't about being a code wizard or a math genius. It's more like being a translator, guiding the conversation between humans and AI to make sure they understand each other perfectly. Because it really is a field where a writer or a research assistant can shine just as brightly as a coder. Because as you might know, a lot of these AI models today are only as good as the inputs you give them. So really, prompt engineering is the art of asking the correct questions. And that is no easy task. So here are three skills that will help you ask better questions that you might not have expected. First up, we have philosophy. Whoa, whoa, whoa, whoa, whoa.
You go. Are you seriously about to turn talk about philosophy inside of an AI related video? How do those two link?
Well, my friend, you might just be surprised because if we go back to the good old days of the ancient Greek thinkers like Plato and Aristotle, we might just uncover some techniques that will help us craft higher quality prompts. Because in the realm of prompt engineering, philosophy really equips you with the ability to think critically, understand complex concepts and apply logic.
But Igor, I'm not trying to solve the meaning of life. I just want to get better results from AI.
Alright, alright, it's time for a practical example. So look, a basic prompt might look like this. What makes a good marketing campaign? Just a simple question that gives you some info. No level of sophistication or depth here, right? So now let's add a philosophical touch to this little prompt here. Because this prompt right here tells us, well, you have to have a strong value proposition. And you know, you should be consistent and use analytics. Classic basic GPT4 response. Nothing groundbreaking here, but what if we spiced things up with a bit of Aristotle? That's right, our Greek buddy Aristotle gave us the concept of each ethos, pathos and logos. Those are the three pillars of persuasion. And if we apply these concepts here, it could look something like this. Discuss the elements of a successful marketing campaign. Same as before, considering Aristotle's concept of ethos, pathos and logos. And look, the proposition I'm making here is you gotta Remix your basic prompts with unexpected concepts. And philosophy is full of those. And they're high quality because a lot of people thought a lot about these. And now you're about to reap the rewards here. So have a look at this. They're often applied in marketing to design effective campaigns. And here we go. Finally, some worthy topics. Credibility or ethical appeal. Then we talk about the emotional appeal in a pathos driven campaign. Marketers might use emotionally charged imagery, narratives, music, or other elements that resonate with the target audience or an emotional level that might not be the solution to all your marketing problems. But this is a lot more concrete because we added something more concrete, a concept from Aristotle. A. All right, let me give you another example here. So again, basic prompt, how can I write better sales copy? We run this, we get a very general answer. I should understand my target audience and I should have a call to action and tell a story because that's engaging. All right, all true, but this is not what you're looking for if you're not a beginner. And now the touch of philosophy, let's just extend it with based on the Socratic method of questioning to understand customer needs better. All right, as you can see, we got a whole lot more concrete there. We're focusing on one topic to get precise results for that and we're using the Socratic method of questioning to explore that specific topic. So we have very specific instructions, very specific context, and a very specific method to explore that that results in a very specific answer that is actually useful and not just for beginners and, I don't know, judge yourself. Is this any better? Is this more useful? Like this just produced a step by step process on how to really understand your customers. It starts generally, but then you uncover pain points, implement solutions, position yourself as the solution, then you show the benefits and overcome objections. It's a great little framework that you can now apply to any topic. And now you can follow up with write me a sample for my prompt engineering course and it will do just that. You will get sales copy in this structure. And there you go. Now that I look at it, I would need a little more context as to what prompt engineering is in the context of ChatGPT, as this goes, a little too broad, but you will only learn that by trying. So as always, we'll just make it more specific. Not bad. Wouldn't it be a game changer to have such a comprehensive, flexible and interactive approach to learning engineering, presenting our all in one prompt engineering course? But again, the point here is that philosophy allows you to go deep, because the deeper you go, the more specific you're getting. And specific inputs equal specific outputs. Now, if you want to start exploring philosophy, there's one podcast that I would recommend. It is called Philosophize this and I personally listen to over 100 episodes. But then I do find listening easier than reading, so that's why I took that route. I would like to introduce today's sponsor, Brilliant.org, which is a platform that makes learning feel like you're embarking on a thrilling quest. It's like a real life game of knowledge where you always win. Brilliant.org is your best buddy when it comes to conquering the realms of math, computer science in a super interactive way. Just imagine this for a second. You're unraveling complex ideas, but instead of being a drag, you're actually having a ton of fun. And that's possible because Brilliant gives learning this gaming twist with a whole bunch of captivating challenges, friendly rivalries. It's like exploring a universe of wisdom with fellow knowledge hunters and loving every minute of it. A whole different story than just reading a textbook, right? And guess what? I've got a little treat for you. You can check out the awesomeness that is brilliant.org absolutely for free for a whole month. Just head over to brilliant.org daidvantage or click the link in the description. And the first 200 of you that head on over to that Link will get 20% off the annual premium subscription. So why not give it a try? It's totally free for a month and you might just pick up a digital skill that will benefit you throughout the rest of your life. And now let's get back to the video. But next up, we gotta talk about storytelling. Cause this skill goes a long way when it comes to prompt engineering. It's really about crafting a compelling narrative that resonates emotionally with your audience and provides context to enable them to think and engage more deeply. One of the all time greats in the discipline of storytelling was Steve Jobs. I don't think the man needs an introduction, but you gotta consider that his skills in storytelling went all the way from Apple to Pixar, back to Apple, and they changed the game in all of those companies. He was really known for his ability to weave captivating narratives around the products that he was developing. And he transformed technical launches that are usually boring and just for the nerds into a spectacle where people tuned in to check out what he had to say. And that's all for the power of story. This man didn't just sell products, he sold experiences and visions of the future. And nowadays it's one of the most valuable brands on planet Earth. So there's probably something there. Now let's apply some of his techniques in the context of prompt engineering. So let's start with something basic again. Write a marketing strategy for our product. No need to show you how basic this is anymore. Let me jump straight to the storytelling angle, to this prompt. There you go. Imagine you're a marketing executive at a startup launching a new fitness app that uses AI to create a personalized workout plan. Write a blog post introducing Sarah, a busy working mom struggling to find time to exercise and show how our app can help her overcome her challenges and achieve her fitness goals. Now this approach I've been experimenting a lot with lately because it produces very different results. If you talk to GPT4 in this storytelling fashion, you just get a different angle to the result. Have a look for yourself as per usual, different input, different output. And guess what? If you input a story, it's going to come out with a story. Life is all about balancing the spinning plates of family, work and personal well being. No one knows this better than Sarah, a hard working mom juggling a full time job, a bustling family life and her own fitness goals. As you can see, directly took the storytelling approach, which actually for content creation often is the key to unlocking more engagement. And as I always say, the number one way to get better results from ChatGPT is to include more context that is relevant. There's no easier way to do that than to simply tell a story because that's how we would communicate with other people. Might as well leverage that skill set to communicate with AI. The result is more persuasive and relatable. And all of a sudden GPT4 is not just filling in the gaps with various points of information, it's telling engaging story where everything ties into each other. And if you want to get better at storytelling, there's this one book I would recommend called Building a Story Brand and it's more of a marketing book, but it's one of the most relevant resources that I found on how to learn storytelling and how to apply it to your writing or your prompts. However, there's one more skill I need you to know about when it comes to prompt engineering which you might not have expected and that is the skill of creativity. Because as you can see, being a good prompt engineer is not just about having technical knowledge, it's a part of it, but the main part really is using those technical understandings to creatively solve problems. Because trust me when I tell you that creativity is way more important in this discipline than most people imagine. As crafting the right prompt is often an iterative process where you run into a wall 10, 15, 20 times and then you find an angle that works and gets the results you want. Well, going through all iterations take some creativity. I think it's time for an example. Okay, so how about this? Describe a day at the beach. A solid one out of 10 when it comes to prompt quality. I mean, seriously, you're giving GPT 4. Nothing to work with here. Just like invent everything. The character, the weather, the story, everything. Just a day at the beach doesn't get more generic than this. I mean, I guess you could say describe a day. Okay, this is a 0 out of 10. But how about starting with the keyword imagine? Because if you start with imagine, it opens up a lot of doors because you're not tied to reality anymore, right? You can pretty much imagine everything. So. So let's say imagine you're a sand grain at a busy beach. Now describe your day. That's pretty creative. A completely different angle than we just imagined. A scenario that wouldn't occur to us in our day to day. And guess what? Creative in creative out. Early morning surfers appear. Their board presses down on my companions and me before they head out into the pulsating ocean. As the surf crashes onto the shore, I'm dampened by the salty sea spray that washes over us before the sun once again dries me out. That is pretty damn creative. Okay, but instead of looking at the day in the life of a sand grain on the beach, let's look at something a little more practical. Discuss the importance of social media marketing. A classic prompt that I would rate at a 1 out of 10 doesn't mean the result is a 1 out of 10. GPT4 is smart, but you probably know what I mean. This is just going to produce beginner friendly results. Now what if we add a little bit of that creativity? Check this out. What if I say imagine you're a brand new social media platform that's trying to compete with TikTok and Instagram. How would you convince users to create more content on your platform? Now this is a completely different approach, right? As we're going backwards from the results that they're looking for here, right? We don't even care if it's a new social platform. We're using creativity to look at what strategies these platforms might be deploying. To attract more users away from their competition, AKA strategies. And if you align yourself with those strategies, you might just do better on these platforms. Right? Let's have a look at the result. All right, so if you read this, you'll find out that they're going to need some ways for people to collaborate. And guess what? If they provide either and you participate, they're going to be pushing you a little more. Right? Many other points here that you could use to your advantage to understand how an entity like this would think. And we got these results by asking a non obvious question by deploying a little bit of creativity. So look, if you want to get better at these skills, it's not just about learning some of the technical concepts. It can also be about educating yourself in some of these unexpected skills. But let's not forget other skills that are essential here too. Strong communication, critical thinking, attention to detail, and as mentioned, a deep understanding of your model and how it works. But that comes mostly from experience. And if you found this video helpful and you want to learn more, well, you would probably enjoy my Prompt Engineering course where we explore various topics that enhance your ability to craft better prompts. The first chapter that gets you set up is free. And if you want more prompt engineering tips on YouTube right now, well, this video will show you exactly what mistakes to avoid when crafting prompts. I'll see you there.










