I have been blogging since 2010. Sixteen years. I wrote hundreds of blog posts before I ever had access to generative AI. I developed opinions, changed theological positions, explored philosophy, wrestled with consciousness and spirituality, and found my own way of expressing myself long before ChatGPT showed up. So when an AI detector announces that something I wrote is “70% AI-generated,” forgive me if I am not terribly impressed.
What exactly has it detected? Did it detect that I suddenly
became capable of writing coherent paragraphs? I was doing that before AI. Did
it detect my vocabulary, my sentence structure, the way I develop an argument,
or the subjects I return to repeatedly? Those existed before AI too. This is
one reason I think the growing obsession with detecting AI writing is becoming
ridiculous. We have somehow decided that if software believes AI participated
in producing a piece of writing, we have discovered something important about
its intellectual legitimacy. We haven't.
The question that matters is not whether AI helped write
something. The question is whose ideas are being expressed. I use AI
unapologetically. I use it for research. I use it to explore ideas, find
thinkers I haven't encountered, compare philosophical and theological
positions, challenge my conclusions, and organize thoughts that are sometimes
bouncing around in my head faster than I can organize them myself. And yes, I
use it to help turn those thoughts into readable prose. I consider that
intelligent use of technology.
There is an enormous difference between using AI because you
have nothing to say and using AI because you have so much to say that you have
discovered an extraordinarily powerful tool for helping you say it. That
distinction seems to get lost in the current hysteria over AI writing.
If someone knows nothing about consciousness, asks AI to
write a thousand words about consciousness, doesn't understand the result, and
publishes it under his name as though he were an authority, I understand the
criticism. But suppose someone has spent years thinking about consciousness.
Suppose he has read widely, developed positions, rejected other positions,
changed his mind repeatedly, and built an interconnected worldview involving
science, philosophy, theology, mysticism, psychology, and personal experience.
Then suppose he gives AI that intellectual framework and says, “Help me develop
this idea.” That is something entirely different. AI didn't create the
worldview. AI entered into a worldview that was already there.
In my particular case, there is a sixteen-year paper trail.
Hundreds of blog posts exist from the years before I ever used generative AI.
If someone genuinely wants to know whether the ideas I publish today belong to
me, they don't need an AI detector. They can read what I was writing ten or
fifteen years ago. They can watch ideas emerge, develop, change, and sometimes
disappear entirely. My intellectual fingerprints were already all over the
place before AI arrived.
What AI has changed is my ability to work with those ideas.
I can take a thought that occurs to me over morning coffee and immediately
begin interrogating it. Has anyone else proposed something similar? What would
the strongest objection be? Does this conflict with something else I believe?
Is there scientific research relevant to it? What would a particular
philosopher or theologian say about it? What happens if I connect this idea
with another idea I've been developing? That isn't AI replacing my thinking.
That is AI allowing me to think interactively, and I find that intellectually
exhilarating.
The accusation that AI-assisted writing is somehow
inherently dishonest also rests upon a strangely romantic idea of authorship.
Apparently, the only authentic writer is a solitary individual sitting in front
of a blank screen painstakingly typing every sentence without assistance. But
writers have never really worked that way. Authors have editors. Scholars have
research assistants. Public figures have speechwriters. Publishers restructure
manuscripts. Journalists have copy editors. Writers use dictionaries,
thesauruses, encyclopedias, search engines, grammar software, citation
managers, spell checkers, and word processors. Nobody looks at a beautifully
edited book and says that because an editor changed some sentences, the author
didn't really write it. But introduce artificial intelligence and suddenly we
are conducting an intellectual crime scene investigation.
Then come the AI detectors. A piece of software analyzes
another piece of writing and produces an authoritative-looking percentage: 47
percent AI, 68 percent AI, 82 percent AI. And what precisely does that
percentage establish? It doesn't tell you who conceived the argument. It
doesn't tell you who selected the subject. It doesn't tell you whether the
author understands what was written or agrees with every sentence. It doesn't
tell you how many times the text was revised. It certainly doesn't tell you whether
AI generated the original idea or merely helped articulate something the writer
has been thinking about for twenty years. Most importantly, it cannot crawl
inside the writer's head and determine intellectual ownership. Yet somehow we
have begun treating these percentages almost like laboratory results.
Of course AI can be abused. Students can use it to avoid
learning. People can publish material they don't understand. AI can make
mistakes, fabricate information, invent citations, and occasionally produce
confident nonsense. Anyone using it seriously has a responsibility to check
important claims and take responsibility for the final product. But misuse of a
tool is not an argument against the tool.
I don't surrender my judgment to AI. If it gives me
something I don't believe, I reject it. If it misunderstands my theology,
philosophy, or view of consciousness, I correct it. If it misrepresents my
position, I change it. If it produces language that doesn't sound like
something I would say, I tell it so. That process itself requires thought.
Ultimately, I decide what gets published under my name, and I accept
responsibility for it.
That is why I think we need to reconsider what authorship
actually means in the age of artificial intelligence. Authorship isn't
fundamentally about keystrokes. It is about intellectual ownership. If I can
sit across the table from you and explain what I wrote, defend it, elaborate
upon it, answer objections to it, tell you where the idea came from, and
explain how it fits into the larger body of things I have been thinking and
writing about for years, then telling me the idea isn't really mine because AI
helped construct some of the sentences strikes me as absurd.
In my case there is an additional irony. AI can learn to
write more like me precisely because I have spent years developing a
recognizable body of thought and writing. I bring the ideas. I bring the
questions. I bring the history. I bring the theological evolution,
philosophical curiosity, contradictions, experiences, intuitions, and
occasionally a crazy hypothesis that I want to chase down simply to see where
it leads. AI brings access to information, organizational ability, comparison,
pattern recognition, and linguistic fluency. I don't see that as replacement. I
see it as collaboration.
I suspect this is where writing is headed whether the
AI-detection industry likes it or not. So yes, I use AI to write. I use it
enthusiastically. I am not embarrassed by it, I am not trying to hide it, and I
don't need a piece of software assigning percentages to my paragraphs to tell
me whether the ideas in them belong to me. I know where the ideas came from.
The real test shouldn't be whether AI helped me write
something. Ask whether I actually believe what I published. Ask whether I
understand it. Ask whether I can explain it, defend it, and expand upon it. Ask
whether there is actually something of me in what you are reading. Those are
meaningful questions about authorship.
And if the answer is yes, I have no hesitation whatsoever
putting my name at the bottom.
AI may have helped me write it. But I am the one who had
something to say.







