What Daily AI Use Actually Taught Me About Leading With It
What does human and AI leadership actually look like in practice?
It looks like supervising the tool instead of trusting it on autopilot. In practice that means catching the assumptions AI makes without being told, using it to get past the blank page instead of skipping your own thinking, and keeping judgment, values, and client relationships where they belong, both at the center of everything and in the hands of humans.
We believe that human and AI leadership will be the next big shift in how businesses run. There is data behind the buzz too. Leaders and managers now use generative AI several times a week at a rate far ahead of their frontline teams, a gap researchers have started calling the silicon ceiling. More than half of U.S. employees now use some form of AI in their role, a milestone Gallup’s polling crossed for the first time this year.
I have not been following this trend from the outside, but instead I have been living inside it, testing these tools every day at Triple Creeks Consulting. That is how we operate here with every tech tool. If we would not use a tool ourselves first, we do not put it in front of a client. And the deeper I get into daily use, the more I notice the limitations, the biases, and the quiet tendency these tools have to assume.
What does human and AI leadership actually look like day to day?
Most days it is quality control. I rarely start from an empty page anymore, which used to be my biggest creative block in this work, and AI gets me moving which always feels great.
But moving in a direction is not necessarily the same as moving in the right direction. The leadership part of this practice is staying close enough to the work to catch it when the direction drifts, and knowing where the parts of each draft came from.
I would actually call myself a fairly creative person, and using AI daily has sharpened that rather than replaced it. I am learning how differently people use these tools, every day. Some use them for structure, some for brainstorming, some purely for editing. And the more I see that range, the more I realize there is no single ‘right’ way to use AI, but there is the discipline of staying involved in your own work.
This is also where it gets easy to pigeonhole yourself. It is tempting to frame every use of AI as outsourcing, handing off your creativity or your judgment because the tool can produce something fast. The more honest framing, at least in my experience, is that AI works best when it complements a skill you already have instead of replacing the need to have it in the first place.
Why does AI keep assuming things you never told it?
I gave one of my AI tools a standing instruction in its memory settings telling it not to assume anything about my requests. Almost immediately it started asking me a long list of clarifying questions on nearly every single prompt. To be honest, it was super annoying.
That reaction told me something important: all that questioning had been happening the entire time before I gave the instruction, of course, it was just invisible. The assumptions were baked into every answer I got, and I never saw them because the tool never had instructions to surface them. Eventually, I got used to the annoying questions and my Claude still isn’t supposed to assume anything.
Researchers are documenting the same pattern at scale. A Microsoft and Carnegie Mellon study of knowledge workers found that the more confidence people had in an AI tool’s ability to complete a task, the less critical thinking they actually applied while using it. Some researchers now call this cognitive surrender, where workers accept AI output with barely a glance instead of treating it as a first draft that still needs a human read. This is the opposite of leading with AI.
Real leadership here means staying the person who keeps questioning the output, not the person who stops asking questions because the tool sounds confident.
Where do humans and AI actually split the work well?
My team is genuinely good at a lot of the same things AI is good at. We are strong at implementing operational systems once we know the direction. Give either my team or an AI tool clear instructions and a defined process, and both can execute well. That overlap is exactly why our adoption and coaching around it requires more intention than the average.
Research on which activities resist automation points to the same divide. Tasks that involve managing and developing people, or applying expert judgment to open-ended decisions remain far harder to hand off than repetitive operational work.
So the real skill is deciding what stays yours, intentionally and with a purpose. In our own work, that usually means AI handles drafts, structure, and first passes, while judgment calls about a specific client’s situation, their team dynamics, and their values, stay with a person in the room. The tool speeds up the parts that were always mechanical and empower your team with tools to do their jobs well.
What does the real cost of keeping up actually look like?
Part of the pressure behind the whole human and AI leadership conversation is speed. Entire industries now expect faster output because AI made faster possible, whether or not that pace is actually healthy for the people doing the work. That snowball effect is real.
What tends to get lost in the speed conversation is cost. A single AI query is small on its own, using roughly the same energy as a typical web search according to recent independent analysis. But scale changes that math fast. Some organizations are now running through thousands of dollars a day in AI usage, and most people using these tools daily have very little sense of what that actually means in energy, water, or dollars behind the interface they are typing into.
Keeping up with AI in a human way means asking that cost benefit question honestly instead of assuming faster is automatically better for your team or your mission. Before you add another tool or another daily habit built around AI, it is worth asking what you are actually trading for that speed, and whether your team even has the capacity to absorb a faster pace right now.
What this actually changes for your leadership
The version of human and AI leadership that actually holds up is not about outsourcing your thinking or your creativity to a tool; it is about using AI to explore more of what you already have to offer, and then protecting the time that frees up for the work only you can do.
For me, that has meant more time for creative things with my daughter, time I would not otherwise have made room for. That is the trade worth making. A more honest accounting of where your time and attention actually go.
That is the whole trend, stripped of the buzzword. Not a race to adopt every tool first, and not a refusal to touch any of it either. Just leadership that stays awake while the tool works, asks what it assumed, and decides (on purpose and with intention) what still belongs to a person.
If your organization is trying to figure out where AI genuinely belongs in your systems, and where it should never be the one making the call, that is exactly the kind of ethical AI adoption work we help clients build. We start with process and operational structuring so the line between what AI handles and what stays human is clear and intentional from the beginning.
Book a free call and let’s map out where AI should sit in your organization, and where your judgment needs to stay firmly in charge. Schedule a call with Triple Creeks.
