The data is in, and the results are shocking. My latest experiment in AI-powered filmmaking generated a 44% lift in engagement with a fraction of the effort. I’m not talking about some incremental improvement. I’m talking about a leap that changes the game for creators.
For years, I’ve been in the trenches with AI. As an entrepreneur, I built and sold two companies, RemoteTeam and MovieLaLa, where we were constantly trying to connect the right content to the right person. As an investor in over 200 startups, including foundational AI companies like Anthropic, OpenAI, and Scale AI, I’ve had a front-row seat to the revolution. But for all that access, I still made every mistake in the book when it came to applying AI to my own creative work.
I’ve seen the power of personalization, and I’ve felt the pain of getting it wrong. This isn’t another high-level think piece about the “potential” of AI. This is my personal playbook, distilled from hundreds of failures and a few massive wins, to help you shortcut your path to success. This is what every creator needs to know.
The Old World of Content
Remember the old way? We’d create one piece of content for everyone. A single blog post, a single video, a single ad. We’d blast it out and hope for the best. At MovieLaLa, we spent countless hours and a small fortune trying to build a recommendation engine that felt personal. We thought we were cutting edge, but we were essentially just making slightly better guesses based on broad categories. We were personalizing, but not hyper-personalizing.
We were creating for the average user. The problem is, the average user doesn’t exist. Your audience is a collection of individuals, each with their own unique tastes, problems, and desires. Creating for the average is creating for no one.
The New Foundation: Hyper-Personalization
Hyper-personalization isn’t just about adding a {{first_name}} tag to your email. It’s about creating content that feels like it was made for an audience of one. It’s about using AI to understand your audience on a deep, almost intimate level, and then crafting a message that speaks directly to them.
This is a fundamental shift. It’s moving from a one-to-many model to a one-to-one model, at scale. And it’s the single biggest opportunity for creators today.
Here’s my blueprint for making it happen.
My Personal Playbook
Step 1: Go Deeper Than Demographics
The first mistake everyone makes is thinking about their audience in terms of broad strokes: age, gender, location. That’s table stakes. To truly personalize, you need to understand psychographics. What are their values? What are their fears? What are they trying to achieve?
When we were building RemoteTeam, we didn’t just target “remote companies.” We targeted founders who were terrified of losing their company culture as they scaled. We targeted managers who were struggling to keep their teams engaged. We understood their specific pain, and we created content that spoke directly to that pain. That’s why Gusto acquired us. We solved a very specific, very painful problem.
Your job is to become a detective. Use surveys, have conversations, and analyze the data you have. What are the common questions people ask? What are the recurring themes in their feedback? This is the raw material for your AI.
Step 2: The Data Is Your Foundation
AI models are powerful, but they are not magic. The quality of your output is directly tied to the quality of your input. Garbage in, garbage out. Simple as that.
I once wasted a month on an AI storytelling project that produced the most generic, soulless content imaginable. The problem? I fed it a diet of generic, soulless articles. I was trying to create something unique from a foundation of mediocrity.
Your best data is your own. Your past articles, your customer interviews, your unique perspective. This is your secret weapon. You need to feed the AI your voice, your stories, your data. That’s how you create something that no one else can.
Step 3: The Model Is Your Tool, Not Your Master
As an investor in OpenAI, Anthropic, and Hugging Face, I get a lot of questions about which model is “best.” It’s the wrong question. It’s like asking a carpenter which hammer is best. It depends on the job.
Some models are great at creative writing. Others excel at data analysis. Some are fast and cheap, others are slow and powerful. Your job isn’t to find the one perfect model. It’s to understand the strengths and weaknesses of the available tools and choose the right one for the task at hand.
Don’t get caught up in the hype. Start with a general-purpose model and experiment. As you get more advanced, you can start to explore more specialized options. The key is to stay focused on the outcome, not the technology.
Step 4: The Prompt Is Your Art
This is where the magic happens. The prompt is the bridge between your intention and the AI’s execution. A well-crafted prompt is the difference between a masterpiece and a mess.
I’ve seen a single word change the entire output of a model. In my 44% lift experiment, the breakthrough came when I stopped asking the AI to “create a video script” and started asking it to “write a script in the style of a Casey Neistat vlog, but for a tech founder who wants to explain the importance of data quality.”
Specificity is everything. Give the AI a role, a voice, a format, and a goal. Tell it who the audience is. Give it examples of what you like and what you don’t like. The more context you provide, the better the result will be.
Step 5: The Relentless Pursuit of Iteration
That 44% lift didn’t happen overnight. It was the result of dozens of experiments. We tested different hooks, different story structures, different calls to action. Most of them failed. But with each failure, we learned something. We fed those learnings back into the process, and we got a little bit better each time.
This is the mindset of a top 1% performer. You don’t expect to get it right on the first try. You embrace the process of iteration. You test, you measure, you learn, you repeat. The AI makes this process faster than ever before. You can test ten different video hooks in the time it used to take to write one.
My Scars: The Pitfalls to Avoid
I’ve made a lot of mistakes on this journey. Here are a few of the biggest ones:
- Losing the Human Touch: In my excitement about the technology, I once let an AI generate an entire email campaign. The open rates were abysmal. It was technically perfect, but it had no soul. The AI is your co-pilot, not the pilot. Your voice, your stories, your personality—that’s what people connect with.
- Ignoring the Story: I’ve seen so many creators get lost in the tech. They talk about the model they used or the prompt they crafted. Nobody cares. They care about the story. The technology is just a tool to help you tell that story more effectively.
- Forgetting the Goal: Why are you creating this content in the first place? What do you want people to do after they watch or read it? If you don’t have a clear answer to that question, you’re just making noise.
The New Reality for Creators
Hyper-personalization isn't just a new tactic; it's the new foundation. It’s the difference between shouting into the void and having a one-on-one conversation with every single person in your audience.
Creators who ignore this shift will become obsolete. It’s that simple. The tools are here, the data is available, and the cost is dropping every day.
The only thing missing is your unique voice, your unique perspective, and your willingness to experiment. The playbook is in your hands. Now go build.
Frequently Asked Questions
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
How has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.