I've been thinking a lot lately about why generative AI has become such a compelling tool for so many of us. I have spent significant time considering my own thoughts and feelings as well as those of my friends. The use of generative AI has exploded in many areas of life, and we are currently living in the midst of an era of AI experimentation. However, what I'd actually like to discuss is the observation that the way many people interact with AI tends to mimic certain aspects of human collaboration, and I am wondering where to draw healthy lines. ## The Simulation of Collaboration There's something uniquely satisfying about sharing a half-formed idea with an AI system and watching it respond. It feels like collaboration: a digital companion that takes your thoughts seriously and helps you explore them further. In my experimentation, I can share a vague notion of an idea related to product development or scientific methodology, the AI responds with questions and suggestions that feel real. The ideas aren't the same as brainstorming with a colleague, but they often get me to have other creative ideas of my own. But why do so many people choose an AI tool to be their collaboration partner? For starters, most current gen AI systems are have a cheerful tone and tend to be open to taking a different approach. As a result AI feels like it is a judgment-free zone. There's no raised eyebrow, no subtle shift in body language suggesting your idea might be ridiculous. The AI doesn't project that it has a reputation to protect or a competing agenda. It simply responds, with what is most statistically logical based on it's training and reinforcement. This reminds me of times in the lab when someone said something seemingly off the wall or asked a naive question that would completely reframe how we approached a problem. The type of feedback where they were just as surprised as you that it mattered. AI can serve a similar function, asking the basic questions that help us clarify our own thinking. ## The Missing Human Element Yet there's something fundamentally missing in AI collaboration. Having spent significant time working to support customers struggling with complex workflows, I've learned that, at the cutting edge, the leaps tend to be larger and often look like unexpected connections that are built on lived experiences. The frustration associated with troubleshooting, the anticipation when waiting to analyze the results, and the joy that comes from solving one problem only to uncover another ten. No amount of text from a computer could prepare you for this type of work. AI lacks the lived experience that allows humans to make those intuitive leaps—connecting, for instance, a pattern seen in cell behavior to an unrelated phenomenon in customer adoption. It can only draw connections within its training data, missing the implicit, rarely articulated insights that often lead to breakthroughs that reside at the cutting edge. When developing medical devices, some of our most valuable insights come from having lunch with the frustrated pathologist, asking them about their perspective then returning to the lab to watch them improvise a solution. Likewise great insights have come from hearing a lab technician mention an off-hand observation that didn't seem important to them but connected perfectly with a problem our development team are trying to solve. I sometimes wonder if our embrace of AI collaboration reflects a broader social trend. In increasingly remote work environments, has AI become popular because it's a natural extension of our retreat into ourselves? Is it easier to bounce ideas off a digital entity than to schedule yet another video call? Or perhaps it speaks to a deeper insecurity: being vulnerable, afraid that our ideas aren't good enough to share with colleagues without some preliminary validation. ## Finding Balance Despite these concerns, I've found AI to be an invaluable tool, when used in a balanced and thoughtful way. Here are some approaches I've found effective: 1. **Start with your own ideas.** It is so important to not let our creativity muscles atrophy. Spend time building a system for your creativity. A friend of mine, Mike Schmitz, shares his take on a creativity system, which he calls [the Creativity Flywheel](https://practicalpkm.kit.com/posts/creativity-flywheel). Critical elements of a good system include: capturing your ideas, spending some time curating and iterating on that information, from that information you make disparate connections and then use those connections to create. 2. **Use AI to make questions, not outputs.** Like a good brainstorming session, some of the best insights come from asking the right question. I find getting AI to ask me questions about my ideas can help me to clarify my own thoughts. 3. **Ask meta-questions.** Some of my most productive AI interactions happen when I ask, "What themes connect these ideas?" or "What perspective might I be missing here?" These questions help me step back and see my own thinking more clearly. Do this with AI, but better yet, talk with the people in your life! 4. **Maintain your voice.** As a native english speaker I don't need AI to help me write in english. However, I have heard that AI can help people write more naturally in their second language. That said, when using AI for writing assistance, heavy editing is essential. The goal isn't to have the AI write for you, but to help you express your own thoughts more clearly. Do not loose your voice! 5. **Be BOLD and step away from the screen.** This was a challenge before AI, but it is rewarding to collaborate with other people. Your AI won't get excited when you publish your work. However, nothing beats a team celebration after solving a hard problem together. Remember that some of my best insights still come from casual conversations with colleagues, customer site visits, or simply taking a walk. The allure of AI collaboration is understandable and in this current moment, if you aren't using AI, you may well even feel behind the times. However, as we embrace these tools, it is critical to be mindful not to lose our own agency to think, have ideas, be vulnerable and collaborate with others. It is what makes us human.