I’ve been thinking a lot lately about what it actually means to work with AI.
Not use AI. Not deploy AI. Not “transform the enterprise with AI,” which sounds great in a deck and somehow less great when you have to explain the actual plan.
I mean work with AI, because that’s where things get real. AI is in our inboxes, workflows, meeting notes, dashboards, personal lives, and, let’s be honest, probably in half the content we scroll past online.
And while we spend a lot of time asking how powerful AI can become, I think the better question is this: Are we becoming the kind of humans who can use it well?
That question was at the center of a recent conversation I had on the Alter Everything podcast with Tiankai Feng, Director of Data and AI Strategy at Thoughtworks. The more I’ve thought about the conversation, the more I believe this is one of the biggest topics for analysts, leaders, and data teams right now.
AI does not remove the human. It reveals the human.
One thing Tiankai said really stuck with me: We talk about AI like it has its own agency, but in reality, we built it and direct it. We decide where it fits, what data goes in, what guardrails exist, and what outcomes we accept. That’s a big deal.
It means when AI produces something wrong, risky, biased, misleading, or just plain weird, the question is not only “What did the model do?” but also, “What did we fail to design around it?”
This matters deeply for analysts.
Analysts have always lived in the space between data and decisions. You take messy inputs, apply judgment, build workflows, test assumptions, and help organizations act with more confidence. AI changes the speed and scale of that work, but it does not eliminate the need for judgment. I think it makes judgment more important.
AI can summarize faster and generate options faster than we can. It can help build workflows, write formulas, classify data, and suggest next steps. Cool. Great. I’m here for that.
But AI cannot know what matters to your business unless you give it the right context. It cannot decide what level of risk is acceptable or understand the political, operational, financial, and human realities of a decision unless humans design those realities into the process.
That’s where analysts become more valuable, not less.
The analyst of the future is not just a builder of reports or workflows. The analyst becomes a translator, validator, strategist, and steward of how AI-supported decisions get made.
The useful question to ask before using AI
One of my favorite parts of the conversation with Tiankai was this idea that we need to define our relationship with AI.
Sometimes AI is an assistant. Sometimes it’s an intern. Sometimes it’s a collaborator. And sometimes it’s just a calculator with better manners. That relationship should change depending on the task.
If I’m asking AI to summarize meeting notes, fine. If I’m asking it to help me brainstorm five ways to explain a technical concept, great. If I’m asking it to move millions of dollars between accounts, make a hiring decision, approve a legal contract, or make a decision that affects someone’s life, we better slow down and ask who is accountable.
This is where I think a lot of organizations will either win or get themselves into trouble.
The question “Can AI do this?” is becoming less useful by the day because the answer is increasingly “kind of, maybe, sometimes, depending on how brave you are.”
The better question is: “Should AI do this, and under what conditions?”
For analysts, that means thinking in terms of repeatable, reversible, and auditable work. If a task is repeatable, AI can probably help. If it is reversible, the risk is lower. If it is auditable, humans can understand what happened and why.
We need to protect the human voice at work
There’s another side of this that I didn’t expect to care about as much as I do. AI is changing how we sound to each other.
We’ve all received the email that sounds like it was written by a very polite robot wearing a blazer. It’s technically fine. It checks all the boxes. It might even be useful. But something feels off.
At work, trust is not built only through accuracy. It’s built through tone, intent, personality, and consistency. When everything starts to sound the same, we lose some of the human signal that helps teams work together.
Now, I don’t think using AI to improve communication is cheating. In fact, I think it can be a fantastic use case. If AI helps someone make their message clearer, kinder, sharper, or more useful, awesome. That’s a win.
But if AI strips away the actual person, we have a problem.
I want to read something from a teammate and feel like there is a human being behind it. I want the quirky phrase, the clear opinion, the “hey, this might be a bad idea, but hear me out.” I want the stuff that makes collaboration real.
That matters for analysts, too.
Because analytics is about getting people to trust the answer enough to act. If every explanation, recommendation, and insight sounds generic, we may become more productive while becoming less persuasive.
Critical thinking becomes the non-negotiable skill
There was a moment in the podcast when we talked about deepfakes, misinformation, hallucinations, and all the strange new ways AI can make false things look true.
Inside organizations, bad information has always been expensive. AI just makes it easier to produce bad information at scale, with confidence, and in a format that looks polished enough to believe.
That means critical thinking becomes one of the most valuable skills in the AI era. Can you question the output, trace the data, or test the assumption? That’s analyst work.
And it’s going to matter even more as AI moves from answering questions to taking actions. That is why governance cannot be a bolt-on. It has to be part of how teams work.
The future analyst is more human, not less
So, how do we be human in the age of AI?
I don’t think the answer is to reject AI. But I also don’t think the answer is to hand everything over and hope for the best.
The future belongs to people who can do both: embrace the technology and stay deeply human in how they use it.
Most of all, it means remembering that AI is not the point. The point is better decisions. Better workflows. Better outcomes. Better ways for people and organizations to understand what’s happening and take action.
That’s the work.
And in the age of AI, that work is still very, very human.