AI, Authorship, and the Myth of the “Easy Way Out”
Artificial intelligence has sparked an uncomfortable question: if AI helps produce great writing, who deserves the credit?
Beneath that question lies an even deeper one:
What does it actually mean to think?
There’s a growing assumption—sometimes stated outright, sometimes implied—that using AI to write is taking the easy way out. That if something is well-structured, quickly produced, or rhetorically sharp, then the thinking behind it must have been outsourced.
I don’t think that assumption holds up.
But I do understand where it comes from.
Because something real has changed. Just not what people think.
The Bottleneck Was Never Thinking
For years—decades, really—I’ve had ideas.
Fully formed in some cases, half-formed in others. Ideas about leadership, systems, organizations, human behavior. The kind of ideas you turn over in your head while driving, walking, sitting in meetings, or lying awake at night.
But there was always a bottleneck: expression.
I could record voice memos. I could jot notes. Occasionally I’d write something more formal. But most of it stayed trapped—unstructured, unshared, and ultimately unused.
Not because the thinking wasn’t there.
Because turning thought into clear, written communication is work. Slow, effortful work that is easy to defer.
So the ideas accumulated. Quietly.
AI didn’t create those ideas.
It removed the bottleneck.
What AI Actually Does
When I take a voice memo and run it through AI, I’m not asking:
“What should I think?”
I’m asking:
“Help me say what I already think—clearly, coherently, and in a way others can engage with.”
That’s not new.
Throughout history, people have relied on editors, secretaries, collaborators, and trusted reviewers to help turn ideas into polished communication. AI can now perform some of those supporting functions much more quickly.
But the central responsibility remains the same:
I generate the ideas.
AI helps organize, structure, challenge, and articulate them.
What’s different is speed—and capability.
The greatest change, however, isn’t that AI thinks for me—it’s that it removes much of the friction between ideas and expression.
The tool now also helps:
organize arguments
identify gaps
suggest structure
surface where something doesn’t land
That’s not replacing thinking.
It’s accelerating the parts of expression that used to slow thinking down.
From Private Thinking to Public Output
In my case, what changed wasn’t the thinking.
It was the throughput.
Ideas that would have stayed internal now become:
structured essays
clear arguments
shareable content
And they come out faster than people expect.
So it looks like:
“Suddenly, this person has a lot to say.”
But the reality is:
“This person has had a lot to say for a long time, but now there’s a way to say it.”
AI didn’t generate insight.
It converted latent thinking into visible output.
The Loss of Informal Thinking Space
Part of what changed for me wasn’t just the availability of AI. It was the disappearance of a certain kind of conversation.
I’ve always done my best thinking out loud—through debate, through testing ideas at the edges, often using sarcasm, exaggeration, or playing devil’s advocate to probe a system.
In the right environment, that works.
With people who know you well, there’s trust. They understand your intent. If something lands wrong, you can correct it immediately.
That kind of interaction depends on proximity and context.
And in the last few years, lot of that went away, as a result of COVID, hybrid work and more on-line collaboration.
Conversations became more structured, more scheduled, less exploratory. The informal back-and-forth that allowed ideas to develop gradually became rarer.
That created a new kind of friction—not in thinking, but in expressing that thinking.
Because in writing:
tone is harder to convey
sarcasm is easier to miss
intent is easier to misinterpret
So the risk goes up.
“If I say this the way I would in conversation, will it land wrong?”
That anxiety became part of the bottleneck.
AI as a First Audience
AI, in an unexpected way, helped restore part of that process.
It provides a kind of first audience:
one that reacts to structure
exposes where arguments don’t land
highlights where tone or sarcasm doesn’t carry
Not perfectly. Not with human nuance.
But enough.
Sometimes it doesn’t get the humor. Sometimes it misses the intent.
That’s useful.
Because it forces a question:
“If this doesn’t come through here, will it come through to a broader audience?”
AI doesn’t replace human feedback.
But it recreates the early stage of interaction, the part where ideas are still being shaped, before asking other people to spend time engaging.
Refinement Through Pushback
One of the most powerful aspects of this workflow isn’t just drafting—it’s refinement through challenge.
Before AI, pressure-testing an idea meant:
sharing it with a trusted group
waiting for feedback
iterating slowly
Now, I can ask:
“What are the strongest objections to this?”
“What am I missing?”
“Where would a thoughtful critic push back?”
AI surfaces likely objections quickly.
Not all of them. Not perfectly.
But enough to:
expose obvious gaps
strengthen weak points
clarify assumptions
This isn’t about winning arguments.
It’s about not publishing something that collapses under basic scrutiny.
And there’s an important balance:
The goal isn’t to anticipate every possible objection.
It’s to address the ones that matter most—the ones a fair, thoughtful critic would raise first.
Over-refinement creates something airtight but lifeless.
Good writing leaves room for engagement.
The Line Between Assistance and Substitution
The real distinction in all of this isn’t whether AI is used.
It’s where the thinking lives.
There are two modes:
Expression-driven use: You have ideas. AI helps you articulate them.
Substitution-driven use: You don’t have a clear position. AI generates one.
From the outside, those can look identical.
But internally, they are not.
In one case, AI amplifies thinking.
In the other, it replaces it.
That’s the line that matters.
Effort, Value, and the Illusion of “Too Easy”
Part of the discomfort around AI comes from a deeper assumption: that effort and value should be tightly linked.
If something didn’t take long to produce, it shouldn’t be worth much.
But that’s not how most modern work functions.
When Bill Gates and Microsoft began licensing software, there was backlash. The idea that someone could write code once and be paid for it repeatedly felt wrong to some.
It violated an implicit rule: compensation should match effort.
But the market didn’t adopt that rule.
Because the value wasn’t in the act of copying the software.
It was in:
its usefulness
its reliability
the problems it solved
The same pattern shows up whenever tools increase leverage.
What feels like “too easy” is often just a shift in where the effort lives.
AI doesn’t eliminate effort.
It redistributes it.
Donuts, Systems, and Leverage
Throughout history we’ve confused effort with value because, for most of history, the two were tightly coupled. If you harvested twice as much grain, you worked roughly twice as hard.
Knowledge work has never obeyed that rule. One good idea can be used millions of times.
A simple example is a donut shop. If you make donuts yourself, you’re paid for each one. If you design the system that allows others to make them efficiently, you’re rewarded for the leverage that system creates.
AI doesn’t change this dynamic. It simply lowers the barrier to creating and scaling knowledge.
Leverage and Uneven Outcomes
Leverage has always produced uneven results.
In fields like software, writing, and music, a small number of people capture a disproportionate share of rewards—not necessarily because they worked harder, but because what they created scaled more widely.
This can feel uncomfortable.
It challenges the idea that effort and reward should be closely linked.
And yet, people continue to create.
In the music industry, a few artists achieve massive success, while many skilled artists earn very little.
But they still make music.
Because value to the market and meaning to the creator are not the same thing.
AI doesn’t change this dynamic.
It amplifies it.
It makes it easier to create, easier to distribute, and easier to scale—while also making differences in outcomes more visible.
Rethinking “Winning”
It’s tempting to frame all of this as competition:
stronger arguments
faster responses
better structure
But the goal isn’t to overwhelm.
The most influential ideas aren’t the ones that sound strongest initially.
They’re the ones that hold up over time.
AI can make a weak idea sound good.
It can also help a strong idea become clear.
The difference shows under pressure.
So the goal isn’t to win arguments.
It’s to produce ideas that can withstand scrutiny.
Responsibility in a Leveraged World
Using a more powerful tool doesn’t remove responsibility.
It increases it.
Because it becomes easier to:
sound convincing without being correct
publish quickly without reflecting deeply
adopt ideas that aren’t fully your own
So the standard has to be internal.
A simple version:
If I can’t explain it without AI, I shouldn’t publish it with AI. If I don’t agree with every part of it, it isn’t mine yet.
Effort and value were never the same thing.
But AI makes the gap between them easier to hide.
So the responsibility shifts to the creator to ensure the value is real.
What This Really Is
At its best, this isn’t about automation.
It’s about removing friction:
from refining ideas
from expression
from iteration
It’s about:
getting ideas out of your head
testing them
refining them
sharing them
More consistently. More clearly. At greater scale.
That’s not taking the easy way out.
It’s taking responsibility for making your thinking visible and strong enough to engage with the world.
AI didn’t create my ideas.
It simply removed the bottleneck that kept many of them from becoming shared impact.
Deliberate Use Requires Deliberate Boundaries
Powerful tools deserve intentional use.
Intentional use also means understanding a tool’s capabilities and limitations. AI can hallucinate facts, reinforce biases present in its training, or express unwarranted confidence. Like any powerful professional tool, it should be used with an understanding of where it is reliable, where it is weak, and where human verification remains essential.
Throughout this essay, I’ve described AI as a thinking amplifier, not a thinking substitute. That distinction isn’t automatic. It depends on how the tool is used.
Before beginning a session, I try to decide what role AI will play. Will it help organize ideas? Challenge assumptions? Improve clarity? Act as a first reviewer? Those uses amplify my thinking while leaving the responsibility for judgment and conclusions with me.
Other uses cross a line for me. Asking AI to determine what I believe, generate opinions I haven’t formed, or write arguments I cannot defend shifts from amplification to substitution.
The important decision isn’t made while the AI is responding. It’s made beforehand.
Principles are most valuable when established before we’re under pressure to compromise them.
Powerful tools make it easy to drift beyond our intentions if we haven’t already established our own guardrails.
My boundaries have evolved as both the technology and my understanding have matured, and they’ll likely continue to evolve. But the principle remains the same:
Use AI to strengthen your thinking, not replace it.
Remaining the author of your own thinking also means remaining responsible for verifying facts, exercising judgment, and owning the consequences of what you publish. AI can assist in that process, but it cannot assume that responsibility for you.
Responsible use matters because ideas have value only when they can be shared. Throughout history, countless observations, experiences, and insights have remained locked inside people’s heads because the effort required to express them exceeded the time they had available. Used thoughtfully, AI has the potential to change that. It can help more people contribute their expertise, communicate more clearly, and participate more fully in the exchange of ideas. That possibility is worth pursuing—not by lowering our standards, but by raising our expectations for how we use these tools.
The Bottom Line
AI doesn’t determine authorship.
The user does. And with that authorship comes responsibility.
The question isn’t whether AI was involved. The question is whether the thinking is genuinely yours. Tools have always helped humans express ideas more effectively. Authorship begins long before the first sentence is written—it begins with curiosity, judgment, experience, and the willingness to own both the ideas and their consequences.
It can:
amplify thinking
accelerate expression
strengthen ideas
Or it can:
replace thinking
mask gaps
create the illusion of insight
From the outside, those uses can look the same.
But over time, they diverge.
Because ideas that are truly yours—tested, refined, and honestly held—don’t just sound good.
They hold up.
And no tool can do that part for you.
The goal isn’t to avoid AI.
The goal is to use it in ways that strengthen human thinking, expand human expression, and remain the author of your own ideas.
ROB
Rob Douglas
Founder & Principal, Citadel Professional Solutions
Exploring leadership, systems thinking, and the responsible use of AI to amplify human potential
Comments