Most people think creativity is a gift. It's not. It needs a system. And once you learn that system, you start generating great ideas yourself. You start seeing and connecting things other people miss. You build taste that helps you stand out. And today that matters more than ever. AI can generate thousands of options before you finish your cup of coffee. But more options don't mean you are more original. Used badly, AI will turn you into a commodity faster than you can think. Creativity is not some magical force. The edge you need comes from just two things: your originality and your taste. The two things AI can't generate for us. So today, I'll show you how to create with AI without losing your humanity, your edge. You'll learn how to think more creatively, sharpen your taste, and train your mind to create original ideas. So, let's get started with the story of the strangest band in recent years. AI has created a strange new problem. It can make you invisible faster than ever. In 2025, a new band called the Velvet Sundown became super popular for some reason. They crossed 1 million listeners on Spotify in just over a month. They released two albums and their song reached Spotify's viral 50 chart in Britain, in Norway, in Sweden. And they were the full rock band package: handsome guys, long hair, vintage look, the soulful singer, the whole late 60s rock sound, that vibe. But there was just one small problem: the band did not exist. It was all AI. The voices, the music, the band members were all generated using AI. But their music was so familiar and so generic that it felt like any other band. It was a perfectly acceptable rock band. And that was exactly why it worked. But it is completely generic. Everything looks and feels like everything else. I call that synthetic sameness. Now, synthetic sameness makes the work undifferentiated and inconsequential. Just another drop in the endless flood of frictionless content. And that's the paradox. AI can help you create more. It can help you create faster, but it can also make you completely undifferentiated faster than anything else. That's the very definition of commodity. If everyone has access to something for cheap or free, it stops being an advantage.

You know, I get these two questions a lot. The first one is, am I real or am I AI generated? I am not going to answer that question because what if I said I was real, but I was actually an AI that was trained to say I was real? What a mystery box. Uh, just kidding. Of course I am real. The second question is about editing. Uh I work with this awesome team from an agency called Dragon Fruit Media. They were based in Los Angeles, now in New York. Our editors and our creative directors pour over every visual, every detail. And the other day we were talking about this same topic, creativity and AI. And they talked about this one tool they love, Hicksfield. It is an AI studio to help you make stunning visuals. And this video is sponsored by Hicksfield, which fits perfectly because that's the whole point. Our team genuinely loves the nexus of AI and creativity, and they love using Hicksfield with Claude, but the final choices and final decisions are always made by humans. So the question of standing apart from the crowd comes down to just one thing. Do you know how to push AI far enough that you can get something original out of it? That is where we need the edge framework. The middle is the commodity. Edge is how you climb out of it. E stands for exacting. Do you hold your work to a higher standard? Do you edit out everything that sounds generic and cliche? D stands for differentiated. Does your work move away from the obvious AI defaults? Do you bring your own unique perspective, your own unique point of view? G is grounded. Is it rooted in concrete reality, real world data? Is it verified or is it one of those AI hallucinations that we're so used to now? And the final E is emotional. Does it make someone feel something, something real? Does it come from your heart, from your human emotions? AI can't copy that. Now, let's make it concrete. For example, in one of my favorite movies, The Dark Knight, Christopher Nolan, who is the director of the movie, insisted on flipping an actual massive truck in the middle of the streets of Chicago. Now it would be trivial to fake it using CGI and computer effects and doing it in the real world would take so much extra work. Then why do it? Because Nolan wanted to have the same edge that you and I want. In fact, he was applying the same framework: exacting standards, differentiated approach, grounding it in reality, and all of that created a real emotional impact. And by the way, if you like ideas like these and you want to go deeper, you can subscribe to my newsletter. It's every Tuesday. Link is below. Totally free. So, let's walk through the framework edge one at a time. Starting with the first E, exacting standards. The most dangerous thing about AI creativity is also the most useful thing about it. AI never runs out of versions. If you ask a calculator, what is 8 * 7? You get 56 every time forever. But if you ask an AI to write an opening scene of a film that starts with a sunset over an ocean, it might pick a different scene every single time. Run it five times, you get five different openings. Now, that's not a glitch. That is the fundamental feature of AI. AI lives in the world of probabilities. When you send a prompt, here's what's happening underneath: AI breaks your words into small pieces, they're called tokens. And those tokens are then converted into numbers and they pass through this thing called a neural network, which is just a bunch of equations. And then it tries to understand what you're asking. And when AI responds to your prompt, it predicts what's likely to come next, one word at a time. So if the line was "Once upon a", the AI most likely will pick the word "time". "Once upon a time". Now, lots of other words would fit. "Once upon a rooftop", "once upon a spreadsheet", "once upon a parking lot", but "once upon a time" is more likely because AI has seen it more often and it believes that the probability of that word is higher because it's a very well-known phrase. Now, if you ask a random question, every time you ask, AI is going to pick a different answer because it's just picking from all likely words that can fit. And it's like a constant roll of dice and that's why it can surprise you. For the first time, there is a tool that doesn't just execute your ideas. It can expand them and take them in a direction that you had never imagined before. To me, AI feels like this amazing jazz musician who's playing in a swanky nightclub. You don't know what you're going to get. But here's the problem. A machine that generates endlessly also hands you endless noise. It's called AI slop nowadays. And that's why you need to be exacting because you have to be a ruthless editor. 999 of those thousand options need to be killed until you get that one thing that sings to you. And that is about your judgment, about your taste. Here's an example. A team made a 95-minute AI-generated film that was shown around the Cannes film market just this year in 2026. It was called Hellgrind. Now it was made using the Hicksfield video generation platform and the tools that Hicksfield gives you. It was a video made in 4K, which is sort of high definition, and it was made by 15 people. It took just 14 days and half a million dollars. But here's the thing. Just to make the first 25 minutes of that film, the team generated over 16,000 clips. A machine-generated mountain of material because that's what AI is really good for. But only 253 of those were selected. That's ruthless rejection in action. Let's make it practical. Let's say you are a small business owner who runs a coffee shop and you want to create an ad. Now you can connect a creative tool like Hicksfield to Claude, and you can set it up once. You can go to settings, choose connectors, paste the Hicksfield link and you're done in about 30 seconds. After that, you just talk to Claude in plain language just like you would prompt any AI, and Claude will work with Hicksfield for you. For instance, you can ask Claude to have as many variations of your ad as you want. One is in the warm morning light. Another is like a grainy nostalgic memory. Another has a sleek premium feel to it, and yet another focused on happy customers, you know, laughing, having a good time in your coffee shop. You can have as many variations as you want. Now, 10 years ago, each one of these variations would require a crew that comes into your shop and shoots the footage, and they go back into the editing room for a week, and it would take you thousands and thousands of dollars. Today, AI can give you 10 different directions in one afternoon. That's an incredible gift. A person with an idea and a very small budget can now compete with the largest agencies and studios in the world.

Now we go to the next one: Differentiation. If you're using the same AI the same way everybody else does, how will you stand out? It's not AI's job to make you look different. It's yours. During the 2025 NBA finals, there is a prediction marketing company called Kalshi, and they needed to make an ad for national TV. Now they went to real studios and the studios quoted them six and seven figure budgets and months and months of work. So instead they handed it to one person with AI skills and told him they didn't want anything normal. So he built something completely unhinged: an alien drinking beer, someone swimming in eggs, and a cowboy holding a chihuahua. The company generated 400 trials to then find 15 usable shots. But in about three days and after roughly $2,000, Kalshi had an ad and they aired it during the finals and reached more than 18 million impressions. Now Kalshi could have made a safe bet, but then nobody would have cared. That's the power of differentiation. So differentiation is the real divide. Now AI makes creation easy, but the real competence is in your point of view, your ability to stand out, and that can't come from AI, that has to come from you. Let's make it practical again. Let's walk through an example. Say you are building an app to avoid burnout. Now there are 10 others who are making the same app. They're in the same market and every pitch is going to open with a scene on the beach. So throw out that beach and open on a different feeling instead. Maybe a phone stuck at 1% glowing in the dark. The city is dark. Everyone else is still texting and then the battery runs out. It all goes dark. Same story, but now the room is leaning in because you just didn't say burnout is real. You showed them something. You made them feel something in just 3 seconds. And by the way, that visual was generated with Hicksfield as well. And it was just as easy as writing a prompt. So, you can do this, too. When everyone is making the same point, what's the point of saying it louder? Show it from an angle they haven't seen before. Here's something actionable. Before asking AI to create anything that needs to stand out, ask it three questions: One, what would the most obvious version of this look like? Number two, what are five angles that move away from that default? And number three, which one of those five would make people stop and take notice, and why? That's how you create differentiation. AI can make the obvious version beautiful. Differentiation is finding the angle that others won't.

Now, let's go to the next step. G for grounded. This is where AI's biggest weakness lives. AI makes things up and is always confident about it. The other day I was using Gemini and I pasted a research screenshot and asked it to analyze the data. Now it gave me some insights but as I was interacting with it something felt off. So, I asked a simple follow-up question and I said, "Were you actually able to read the data in the screenshot?" And it said basically, "Sorry, I don't have the ability to read the data." And then I asked, "Okay, so how did you come up with all those insights?" And it said, "I'm sorry, I got ahead of myself and I completely fabricated them." Now, here's the irony. Of course, I know that Gemini does have the ability to read visual information from a screenshot, but suddenly when challenged, it was not just hallucinating facts, it was also confused about its own ability. So, it was like layers and layers of hallucination. And I have seen the same thing with chat GPT too. That is what we forget often about AI. We start treating it like some intelligent machine, some kind of a confident expert. But sometimes AI can slip and fall just like a toddler. That is why grounding matters because AI will often create work that feels weightless but smooth and fluent and impressive but it's not grounded. Nolan could have faked that truck flip in the Dark Knight too, right? And if you use AI to do the same thing right now, this is how it might look. But instead, the team executed a practical stunt with a giant truck on Lasal Street in Chicago. And that sequence works because when you watch it, you feel the physical weight of the real thing in the theater. The street is real. The truck is real. The danger feels real. It's all grounded in reality. So that's the actionable move. Never let AI have the last word on anything that has to be true. You can do three moves to ground it: First, make it show the sources. Ask for the real links, the real numbers, the real data, the real quotes behind it. If it cannot produce one, then that claim is a fiction. The second move is to feed one AI engine to the other, and I do it all the time. I constantly ask Claude to verify Chat GPT's work and ask Chat GPT to verify Gemini's work, and so on. And third, use apps like Google's NotebookLM and feed it the data that's your own. That's what grounds AI like that giant truck in Nolan's movie. The weight has to feel real. And now final E: emotional. The same object can evoke different emotions depending on context. Think about a candle. On a birthday cake it's all about joy. It's about celebration. At a church it's a prayer. During a blackout it's family fun time. But at a vigil it's about grief. And on a date night it's about romance and intimacy. Same flame but it burns a completely different emotion every time. That's the one thing AI cannot do for us. It can generate an image of a phenomenal candle and give it the situational context, even put appropriate music to it, but what emotions will get lit in you when you see that is entirely up to you. Coca-Cola remade their famous holiday ad in 2024 and they used AI, and it had the same feel, the same trucks, the snow, the holiday imagery, the same brand nostalgia and every frame was polished. And yet many people called it soulless. The images were familiar but somehow the emotions didn't land. Now they fixed it in 2025. AI knows how to give you the right images, but only you know what the spirit of Christmas actually feels like. And that's the ultimate takeaway from this whole video. AI today can make anything you want. It makes movies, it makes music, it makes ads, pitches, lessons, whatever you want. But every time something becomes abundant, something else becomes scarce. The parts that come from you having lived your life, your memories, your wins, your losses, the way you think, the way you feel. That's what makes you creative. In the world of AI, that's your edge: you. The machine can paint the candle flame. Only you can make it mean something. Your creation isn't done when it looks right. It's done when it feels right. Thank you and I love