For a long time, I could describe what I was good at relatively easily.
Writing. Research. Content strategy. Taking complicated information and turning it into something people could understand. Coming up with ideas. Organizing information. Looking at a business, an audience, or a problem and figuring out what to communicate and how.
These were skills I had built over years of working in communications and marketing. They were part of how I understood myself professionally.
Then AI came along and started doing a surprising number of them.
It can draft an article in seconds. It can research a topic, summarize a long document, generate campaign ideas, analyze information, reorganize messy notes, rewrite an email, create an outline, and identify patterns in data.
And the better these tools get, the stranger one question becomes:
If AI can do many of the things I thought I was good at, what exactly am I good at now?
When your skills stop feeling as special
I don’t think this question is unique to writers or marketers.
Many of us built professional identities around things that required time, practice, and expertise: writing clearly, researching efficiently, analyzing information, brainstorming ideas, organizing projects, producing reports.
AI doesn’t eliminate the need for these skills. But it changes their scarcity.
Producing a competent first draft feels different when almost anyone can produce one. Summarizing information becomes less distinctive when a tool can do it instantly. Brainstorming twenty ideas isn’t particularly impressive when you can generate fifty in thirty seconds.
But as I’ve used AI more, I’ve noticed something interesting. Some of the abilities I used to consider my main professional strengths are becoming less visible, while others I barely thought of as “skills” are becoming harder to ignore.
The skills hiding underneath the skills
When I ask AI to write something, the important part isn’t necessarily that I could write the first draft myself.
It’s that I can read what it produced and know that something is off.
The argument doesn’t make sense. That sentence sounds impressive but says nothing. We’re answering the wrong question. We’re missing the customer’s actual problem. This research is relevant, but it doesn’t support what we’re trying to say.
Sometimes I can’t articulate the problem immediately. I just know the output isn’t right. Then I have to figure out why.
That ability didn’t appear because of AI. It came from years of writing, editing, communicating, watching campaigns succeed and fail, working across different industries, making mistakes, and learning what good work looks like.
AI only made that layer of the work easier to see.
Pattern recognition. Context. Curiosity. Knowing which question to ask next. Recognizing when something is incomplete. Understanding which details matter and which are noise. Connecting information that initially seems unrelated.
And, increasingly, deciding what is worth doing at all.
AI can give me ten articles, fifty ideas, another email sequence, landing page, report, or workflow. That doesn’t mean any of them should exist.
Someone still has to decide what matters, and what we should prioritize right now.
I’m also discovering a different side of myself
There’s another realization happening alongside this one.
I’ve always had an operational side. I like organizing things, figuring out how the pieces fit together, spotting gaps, solving problems, and improving the way work gets done.
But because I’ve worked in marketing for years, I’ve mostly described myself through marketing disciplines.
The more I experiment with AI, the more I notice that I’m just as interested in the system behind the work as I am in the final output.
Now, when I see a specific gap in a process, I find myself asking: could I build something to help with this?
Maybe it’s an assistant or skill or artifact that helps create product pages more consistently. Maybe it’s a tool supporting a repetitive content workflow. Maybe AI can analyze information that would otherwise take someone hours to review.
I’m becoming interested in identifying a very specific problem, understanding the process around it, and figuring out how AI can fill that gap.
And I don’t think this is an entirely new side of me.
Looking back, I can see it in other parts of my career: organizing information, creating processes, coordinating projects, finding more efficient ways to get things done. AI has given me a new way to act on that instinct.
The output isn’t the whole skill
This has changed the way I think about my career.
I’ve spent years describing myself through disciplines: communications, content marketing, email marketing, strategy. Those descriptions aren’t wrong, but lately they feel incomplete.
When I look across the different things I’ve done, the patterns become clearer.
I like taking messy information and creating structure around it. I like spotting connections. I like figuring out why a process isn’t working and imagining how it could work better. I like translating between different worlds: technical and non-technical, business and audience, ideas and execution.
And more recently, as I’ve started building AI assistants in ChatGPT and Skills in Claude Cowork and experimenting with tools I never imagined myself creating, another pattern has emerged:
I really like building things that solve problems.
For years, I might have described myself primarily as someone who writes, communicates, or creates strategy. AI hasn’t taken those abilities away from me. I use them constantly when working with it.
But it has separated the output from the thinking behind the output.
And once those two things are separated, you start noticing which part you actually enjoy.
What did doing the work teach you to see?
Maybe that’s a more useful question as AI becomes part of more professions.
We’ve been trained to describe ourselves by what we do: writer, designer, analyst, developer, researcher. Increasingly, some of the tasks associated with those identities can be shared with machines.
So I’m becoming more interested in looking underneath the task.
What did years of doing that work teach you to notice? What can you recognize that someone without your experience might miss? What questions do you know to ask? What connections do you make? What do you know not to trust?
When do you recognize that an answer is technically correct but practically useless?
And when AI gives you almost unlimited possibilities, how do you decide which ones are worth pursuing?
I don’t have a clear new professional identity to replace the old one. That’s partly why I’m writing The AI Work Journal. Using AI every day keeps exposing questions I didn’t know I had about work, expertise, creativity, and what I actually want to spend my time doing.
But AI is helping me separate two things I once treated as the same: what I know how to do and what makes me good at what I do.
The first will keep changing as the technology changes.
The second has been accumulating quietly for years.
And maybe one of AI’s unexpected effects is helping us see more clearly what we’ve been bringing to the work, and the way we work, all along.