Don’t Let Slow AI Adoption Cost Your Company

If your organization is dragging its feet on AI adoption while you know it could transform your marketing efforts, don’t wait for permission. Start experimenting with personal side projects that double as solutions to real business challenges.

AI has broken free from corporate containment and now permeates everyday life. My grocery app suggests purchases using AI. Even my dentist recommended an AI-powered tool for better teeth cleaning. The technology is suddenly inescapable.

Skepticism is understandable. AI is evolving at an unprecedented pace. But this isn’t just a shift for email or digital marketing; it’s a societal inflection point.

This moment mirrors the early days of email, when professionals had to teach themselves the technology and build their own systems because companies hadn’t yet embraced it. Once again, individuals hold the power to innovate with emerging tools.

In my previous MarTech column, I noted that early email innovation originated in super-enterprise companies and trickled downward, driven by employees at the largest firms. AI is flipping that dynamic. Innovation is now bubbling up from the bottom, fueled by individuals and mid-to-lower market enterprises exploring AI’s potential.

I know several friends who have built AI-powered apps or websites to solve practical problems. Meanwhile, super-enterprise and enterprise companies often lag behind, constrained by information security concerns, privacy policies, or simple corporate inertia.

How to learn AI right now

If you want to move beyond surface-level AI tasks like copywriting but your company is still debating its use, teach yourself. Don’t wait for your organization to catch up. Instead, prepare yourself for when it finally does.

This advice seems obvious, but it isn’t widely followed. I still encounter marketers who treat AI like a search engine or have little grasp of its capabilities. That means you need deep immersion and hands-on experimentation.

Now is the time to explore and learn independently, without breaching your company’s privacy or security policies. First, understand AI. Then, apply that knowledge to solve practical problems, whether at work or in your personal life. That’s exactly what I did, and it all started in my wine cellar.

I test-drove my AI knowledge on a pressing personal issue: inventorying and managing my 300-bottle wine collection. It may sound frivolous, but I’m a serious enthusiast with a collection built over 20 years. Keeping track of it had been a persistent annoyance.

Learn which tool to use

Understanding each AI tool’s strengths is critical. ChatGPT, Claude, Google Gemini, Microsoft Copilot, and other large language models have distinct capabilities and weaknesses. Claude, for example, is terrible at image generation.

By applying what I learned to a personal chore rather than a business need, I explored each tool thoroughly and later avoided a major professional pitfall: failing because I used the wrong tool.

I needed to inventory my collection without spending hours recording each bottle. So I asked myself, “How can AI help me do this?”

I started with ChatGPT, having already built several dashboards with it. It suggested I photograph each wine label, upload the images, and let pattern recognition build my inventory.

Failure can teach you a lot

That sounded simple. I spent half an hour shooting and uploading pictures. But when I unleashed the app, it failed to recognize many labels and returned incorrect information. It claimed I had a bottle of 1999 Screaming Eagle Cabernet Sauvignon worth $2,985. I wish. Enough errors made me doubt the entire inventory.

I was surprised. ChatGPT had excelled at business forecasting and modeling for my work projects. Why did it fail here? I still don’t know. My guess is I asked it to do something it wasn’t designed for.

I moved to Google’s Gemini, which I’d found useful and responsive before. I gave it the same context and photos. Results were hit-or-miss. Gemini couldn’t process the pictures, and worse, it guessed at labels and usually guessed wrong. Adding context and correcting errors didn’t help. It produced no trustworthy inventory.

My wife, a marketing executive, loves Claude. So I tried it. This time, the results were a pleasant surprise.

Claude recognized most labels, flagged blurry or cropped images, and returned helpful suggestions instead of hallucinations. After I retook the photos it needed, it quickly built a 300-bottle inventory. Exactly what I wanted.

Why did Claude succeed where ChatGPT and Gemini failed?

I used the same context, prompts, and processes with all three. The experiment revealed clear distinctions driven by each model’s infrastructure. Each AI model excelled at different tasks and struggled with others. Does that mean Claude is right for you? Not necessarily.

My wine inventory taught me to understand each system’s unique capabilities and quirks. I’m now applying that knowledge to other projects, selecting the best model for each task.

I know it’s easy to find infographics explaining model differences, but until you experience them yourself, you can’t truly trust someone else’s diagrams.

Knowing these differences gives me an advantage I’ll use when evaluating tech platforms that incorporate AI. You can gain that same edge when your company finally decides on an AI direction, if you’ve already test-driven the technology and understand what fits your needs best.

Building your career with AI knowledge

The education you gain from experimenting with AI platforms will pay off now and later in your career.

According to Litmus’ State of Email 2026 report, AI know-how is the top skillset marketing chiefs seek when hiring. The time and money you invest now in learning AI and experimenting with tools will make you a far more valuable candidate.

For my experiments, I used paid versions for work and home projects. Like signing up for multiple streaming services, the costs add up. Remember, this is an investment in your future.

If you’re just starting, a free or lower-cost tier of each AI model can help you understand platform limits. Once you know that, upgrading to a more expensive tier for the systems you find most useful may make sense.

Some AI users dive deep into coding, hosting Python servers, using GitHub, and getting fully ingrained. Others treat it like a search engine or use it for memes and pictures.

Marketers need to be in the middle. We need to know enough to be useful, to organize and complete tasks with the right AI tool, and to recognize when a human must step in to correct errors.

The only way to reach that point is to start experimenting now. Try all platforms. Make mistakes. Push models to their limits, because knowing those limits is essential.

We are at an inflection point in AI adoption, making now the perfect time to train. It would be nice to do it on the job. But sometimes the easiest way to learn is through a personal project you can apply in your own life.

Your company might be slow to adopt AI. But you don’t have to be.