Artificial Intelligence – From Inspiration to Implementation in 2025

Each year in The Operating Partner, I usually talk about a handful of trends I think will catch on in the following year. But this year, I’m narrowing it down to just one: Artificial Intelligence.

Right now reminds me of the mid-1990s, when I’d come home after high school, head over to Borders, buy an HTML or Photoshop book, and teach myself how to build on the World Wide Web. It was all self-directed learning—just me and my Quadra 630 (and later, my PowerMac G3)—until I finally interned at i33 Communications and worked at Conducive and learned from some of the best.

Back then, it felt like I was an outlier. While my friends were out chasing women or shooting hoops, I was flipping through thousands of pages of books, figuring out everything I could about desktop publishing and the Internet. It wasn’t glamorous, but it was exciting.

That passion eventually turned into a little business. As a high schooler, I started making websites for small businesses in Westchester County. I didn’t know how to run a business, but I figured it out as I went along. Every time I mastered a skill, the tools evolved—what took me hours in Photoshop one day could suddenly be done with a click of a button thanks to a new filter or plugin. It was frustrating but also exhilarating, forcing me to constantly learn and adapt. I even purchased my own first car and had a custom NY State license plate, “WEB DSGN.”? Yes, I was “that” guy.

Looking back, I wouldn’t trade those years for anything. They gave me a baseline understanding of how computers, software, and the web work—a foundation that’s been invaluable ever since.

Fast forward to 2024, and this year has felt strikingly similar to 1995. Replace desktop publishing and the World Wide Web with Artificial Intelligence, and the parallels are uncanny. The same energy, curiosity, and rapid evolution are all here.

But there’s a key difference. The baseline understanding of AI is spreading faster than the web or desktop publishing ever did. Why? Because we now have tools like YouTube, blogs, Udemy, and Meetups—resources that simply didn’t exist back then. In 1995, streaming video over a 28.8kbps modem wasn’t even a dream; today, learning is faster and more accessible than ever.

If you’ve been reading The Operating Partner, you know I’m a big believer in AI and its potential to enhance processes. Sure, there are skeptics—but I’m not one of them. And if you’re with me on this, there’s no better time to jump in than now.

Some of my portfolio company executives are already mandating their teams spend 30 minutes a week “playing” with AI tools. You might scoff—what can you possibly learn in 30 minutes? But over a month, that’s two hours, and for someone who hasn’t touched a single tool, that’s a massive head start.

Back in the ’90s, I didn’t need to be a hardcore tech expert. My strength was understanding how the pieces fit together—the application layer. That’s where the real magic happens, and that’s where I encourage you to focus today. You don’t need to dive deep into the technical minutiae of why one large language model is better than another or obsess over GPU specs. That’ll all change in 12 months anyway. What matters is knowing how to apply the tools to solve real problems.

Start experimenting. Pick a process you want to improve—whether it’s generating content with GenAI, creating AI-based videos, turning blog posts into podcasts with NotebookLM, coding faster with Copilot, or building platforms with no-code tech—and get your hands dirty. No one’s stopping you.

Every experiment you try is a step forward. Each one teaches you something new, giving you an edge in a world that’s moving faster than ever.

This was my playbook in the ’90s. It worked then, and I’m following it again now. I hope you do too. Let's learn a ton together in 2025.

If you are looking for a place to start with AI, check out this LinkedIn post by Conor Grennan as he breaks down his own AI stack.