How I Turned AI Into a Teammate, Not Just a Tool — and How You Can, Too

A few months ago, I was stuck on an email. We were about to announce a delay in releasing a new Jotform product, and I needed to explain the "why" in a way that felt honest, optimistic and clear. Not too corporate. Not too casual. I rewrote the opening five times. Nothing was landing.

Out of frustration more than strategy, I opened up my AI assistant and typed, "Can you help me write a company-wide email about a product delay that's transparent, reassuring and keeps the team motivated for what's next?"

What came back was surprisingly decent. Not publish-ready, but good enough to get me moving again. Then I tried something else. "Rewrite this like you're a product manager talking to power users." Then again. "Now condense it into an email." Suddenly, I wasn't stuck anymore — I had options, tone shifts and plenty of ideas. I wasn't working alone. I had a co-pilot.

AI has changed the way we work, but the truth is, most of us aren't actually using it to its full potential. Here's how to change that.

Related: 7 Ways You Can Use AI to 10x Your Leadership Skills

Large language models (LLMs) like ChatGPT have been in the mainstream for a few years now, and even the most tech-averse among us have dabbled in their offerings.

According to Harvard Business Review, the most popular use of GPTs is "personal and professional support," representing a significant shift from last year, when the reigning category was "technical assistance and troubleshooting." In 2024, users were relying on AI for tasks like generating ideas and looking up the answers to specific questions. Now, their relationship to AI has evolved, with users treating it less like a tool and more like a teammate.

This evolution is key. The more we anthropomorphize AI — not in a sci-fi way, but in a practical, collaborative sense — the more effectively we can leverage it. At this point, I use AI to take a number of tasks off my plate, assigning it roles based on what I need done: If I want to gather stats or summarize articles, I tell it to assume the role of a research assistant. If I want a wireframe copy reviewed, I ask it to act like a UX designer. If I'm refining a pitch or strategy doc, I'll prompt it to behave like an investor or board member, so I can pressure-test my thinking before presenting it to the real audience.

The more intentional you are about how you "cast" your AI assistant, the more value it can deliver — not just as a helper, but as a genuine extension of your thinking.

AI isn't magic — it's responsive. Vague prompts beget vague answers. But with clear direction, including the tone, structure and style you want, you'll unlock much more powerful results.

Say you want to create a blog post to announce a new feature or product. Instead of simply saying, "Write me a blog post," make it specific. "You're the CEO of a tech company who just launched [insert product and description]. Write a 300-500-word blog post that accurately captures the product's features, includes key details, and explains to users how it will improve their day-to-day workflows. The tone should be confident and enthusiastic, and adhere to the company's brand voice [examples below].

I've found it helpful to treat prompt engineering like managing a junior team member. Rather than scrapping the first draft, offer feedback: "Make this less technical," "Add a real-world example," or "Make the first paragraph more engaging." The more specific the prompt and the revisions, the better the outcome.

Related: Why Smart Entrepreneurs Let AI Do the Heavy Business Lifting

One of the biggest missed opportunities of casual AI dabblers is using it sporadically instead of systematically. With just a little bit of upfront time investment, AI has the capacity to streamline nearly every aspect of your daily workflow.

These days, even those without any coding knowledge can easily build customized AI assistants designed to lighten the load of repetitive tasks. From Claude's "Project" to Gemini's "Gem," these platforms enable users to store data like detailed instructions and tone guidelines to keep from having to write the same prompt over and over.

For example, I created a scheduling and email assistant that helps me triage my inbox and prepare meeting follow-ups. It's loaded with previous email templates, understands my preferred tone and can quickly draft context-aware replies or send a recap with action items after a meeting. Because it's trained on how I typically communicate and prioritize, it feels less like a chatbot and more like having a highly-responsive personal assistant.

My advice? Start small. Choose one recurring task — something low-lift but time-consuming — and assign it to your AI assistant. Give it structure, test it out and offer feedback as you go. Before long, these automated assistants will start to feel like invisible team members working behind the scenes to keep you focused on what matters most.