- Tension: We believe AI will either save our careers or destroy them, when the real friction lives in our unwillingness to redefine what “our work” actually means.
- Noise: The productivity discourse drowns out the uncomfortable truth that most content workflows were already broken before AI arrived.
- Direct Message: AI didn’t replace my workflow. It exposed which parts of it were never mine to begin with.
To learn more about our editorial approach, explore The Direct Message methodology.
On January 1st, I made a decision that would have been unthinkable to me a year ago. I handed over my entire content production process to AI tools. Every article outline, every first draft, every social media caption, every email newsletter. For 90 days, I would write nothing from scratch. The experiment came from a place of curiosity mixed with mild panic, the kind that keeps content creators awake at 2 AM wondering if we’re all about to become obsolete.
Three months later, my income is up 34%. I’m publishing twice as much content. And I feel more creatively depleted than I have in years.
This is not the story I expected to tell.
The promise that doesn’t match the experience
When I announced this experiment to my network, the responses split cleanly down the middle. Half cheered me on as a pioneer embracing the inevitable future. The other half warned I was committing professional suicide, that authentic voice cannot be automated, that audiences would smell the artificial mile away.
Both groups shared the same underlying assumption: that AI would either elevate my work to new heights or destroy what made it valuable. What nobody predicted, what I didn’t predict, was that AI would do both simultaneously while revealing something far more uncomfortable about the nature of content work itself.
During my time working with tech companies as a growth strategist, I watched countless businesses chase automation as the solution to scaling challenges. The pattern was always the same: automate the process, hit the efficiency targets, then wonder why customer engagement dropped even as output soared. I thought I understood this dynamic. I didn’t realize I was about to live it.
The expectation was simple: AI handles the grunt work, I focus on strategy and creative direction, income rises because I can produce more without sacrificing quality. The reality proved far stranger. My income did rise, but not because the content got better or even because I produced more of what my audience wanted. It rose because I could flood multiple channels simultaneously, playing a volume game I’d always intellectually rejected.
The hype cycle that obscures the real querstions
The current discourse around AI and creative work operates at two extremes, both equally useless for anyone trying to navigate actual decisions. On one end, the techno-optimists promise that AI will free us from drudgery to focus on “high-value creative work.” On the other, the skeptics warn that we’re automating away the human soul of our professions.
McKinsey estimates that generative AI could add $2.6 to $4.4 trillion annually to the global economy, with marketing and sales among the functions seeing the biggest productivity gains. The promise is seductive: do more with less, scale infinitely, optimize everything.
But here’s what the productivity metrics miss: I was producing content at twice my previous rate, but I couldn’t tell you what 70% of it actually said. I was managing the AI, editing its outputs, directing its focus, but I wasn’t thinking. Not really. The cognitive work shifted from creation to curation, from writing to quality control. I became a supervisor in my own creative process.
The trend cycle demands we choose sides: Are you an AI adopter or a AI skeptic? Are you future-facing or clinging to the past? These binary frameworks prevent us from asking better questions: What parts of the creative process actually benefit from automation? Which parts lose something essential? And what does it mean for our work when the answer is “both at the same time”?
A Harvard Business School and Boston Consulting Group study found that consultants using AI completed tasks 40% higher quality and 25% faster. However, researchers noted the trade-off: while the ideas were superior, they showed reduced variability, suggesting that AI might lead to more homogenized outputs.
That was my 90 days in a sentence.
The uncomfortable truth about automation
Halfway through the experiment, I had a conversation with a long-time reader who said something that stopped me cold: “Your recent articles feel like they’re written by someone who’s read all your old articles.” The comment was meant as constructive feedback. It landed as revelation.
AI didn’t replace my workflow. It exposed which parts of it were never authentically mine to begin with.
The work that AI handles effortlessly is often the work we were already doing on autopilot, following templates, recycling frameworks, optimizing for algorithms over insight.
This is the paradox nobody wants to acknowledge: AI excels at producing exactly the kind of content that dominates our feeds precisely because that content was always somewhat automated, even when humans wrote it. The advice article formula. The listicle structure. The SEO-optimized explainer. We created templates to scale our output, then act surprised when AI executes those templates better than we do.
My income rose because I could produce more of what I’d already been producing. But what I’d been producing was increasingly indistinguishable from what everyone else was producing. AI didn’t make my work generic. It revealed how much of it already was.
Pew Research found that while 58% of Americans have heard of ChatGPT, only 14% have actually used it. The gap between awareness and adoption suggests most people intuitively understand something the productivity evangelists miss: these tools change not just how we work but what our work becomes.
Redefining value in an automated landscape
I’m writing this article by hand. Actually writing it, not prompting an AI and editing the output. It’s taking me four times longer than my AI-assisted articles took. My fingers feel slow on the keyboard. I keep second-guessing sentences I would have let pass in the edit of an AI draft.
And something is happening that didn’t happen during those 90 days: I’m discovering what I think by writing it.
The experiment taught me that AI can replicate my patterns, my structures, my characteristic ways of approaching topics. What it cannot replicate, what I’d stopped doing during those three months, is the messy, uncertain process of figuring out what I actually believe about something by wrestling with how to express it.
The economic reality remains: I can produce more content, reach more people, and generate more income by leveraging AI throughout my workflow. According to Goldman Sachs, generative AI could raise global GDP by 7% over a 10-year period while automating 300 million jobs. Content creation sits squarely in the crosshairs.
But efficiency is not the same as value. Productivity is not the same as purpose. And automation reveals rather than resolves the question every creator must answer: What parts of your work actually require you?
For me, the answer is narrower than I wanted it to be and more important than I’d realized. AI can handle research synthesis, structural organization, even stylistic mimicry. What it cannot do, what I spent 90 days not doing, is bring genuine uncertainty to a topic and discover something through the act of working it out.
The path forward is not rejection of AI or wholesale embrace. Going back to my fully manual workflow feels as absurd as continuing the fully automated one. Instead, I’m learning to treat AI as an amplifier of intention rather than a replacement for thought. Let it handle the research aggregation, the first-pass structuring, the format adaptation across platforms. But return the actual thinking, the real writing, the moments of creative discovery to the human side of the equation.
This means producing less content overall. It means my income will likely stabilize somewhere below that 34% spike. It means accepting that in a content landscape increasingly dominated by high-volume, AI-assisted production, choosing to work more slowly is choosing to be less visible.
But visibility is not the same as significance. And the question that matters is not whether AI can do your work, but whether the work AI can do is the work worth doing.
