What the Declaration of Independence can teach us about AI, originality, and the craft of writing.
I have a confession.
I hate to read.
I hate to write.
At least I hate the act of reading and writing.
I love what they do for me.
Reading and writing are to the mind what jumping jacks and pushups are to the body. Few people enjoy the exercise itself. Most of us enjoy the results. We enjoy being healthier, stronger, and more capable.
The same is true intellectually.
Reading helps me better understand the world around me. Writing helps me better understand myself. Together, they force me to organize thoughts that would otherwise remain vague feelings and half-formed opinions.
That is why, despite my aversion to both activities, I have spent much of my life doing them.
Over the years, I accumulated a library of more than a thousand books, most of them nonfiction. I published technical research papers. I wrote newspaper opinion columns. I wrote a book. More recently, I have written countless blog posts exploring politics, history, and civic life.
None of that makes me special. It simply means I have spent a lot of time doing intellectual jumping jacks and pushups.
Which brings me to ChatGPT.
Some readers have asked whether I use artificial intelligence when writing. The answer is yes.
I use ChatGPT regularly. I use it to research unfamiliar topics, challenge assumptions, identify weaknesses in arguments, improve organization, and help me find the right language for an idea I am struggling to express.
For some people, this raises questions about authorship.
Who really wrote the article?
Who really owns the idea?
Oddly enough, this debate reminds me of Thomas Jefferson.
Jefferson did not invent the ideas contained in the Declaration of Independence. The philosophy of natural rights came largely from John Locke. The grievances against the Crown had circulated throughout the colonies for years. Other declarations and resolutions contained similar language and arguments.
Jefferson acknowledged this reality. He later explained that the Declaration was not intended to introduce new principles. Its purpose was to express what he called the “American mind.”
His contribution was not invention.
It was synthesis.
He gathered ideas, arguments, traditions, and experiences that already existed and arranged them into a coherent whole.
In many ways, that is how I use artificial intelligence.
No, I am not comparing my writing to Jefferson’s. I am comparing the process.
Jefferson relied upon books, letters, conversations, and centuries of accumulated knowledge. I rely upon books, research papers, conversations, and modern digital tools. In both cases, the challenge is not generating information. The challenge is making sense of it.
The critical difference is judgment.
ChatGPT can generate ten different arguments. It cannot decide which argument is true. It cannot determine which evidence is persuasive. It cannot take responsibility for a conclusion.
I must do those things.
The research is still mine.
The thinking is still mine.
The judgment is still mine.
Artificial intelligence simply helps me navigate a larger body of information than I could manage alone.
When I began writing my first book, The Art of the Compromise, I often joked that what I really needed was a research assistant. Better yet, a graduate student willing to spend long hours tracking down sources, checking references, and helping organize mountains of notes. I completed that book before I discovered ChatGPT, and while I am proud of the result, I certainly do not miss some of the drudgery that accompanied the process.
Perhaps that is why I find the criticism of AI-assisted writing somewhat curious. For most of history, authors were valued less for originality than for their ability to gather wisdom, organize ideas, and communicate them effectively. Shakespeare borrowed plots. Jefferson borrowed arguments. Dale Chihuly relies upon teams of skilled assistants to help transform vision into reality. Nearly every great thinker, writer, and creator has stood upon foundations built by others.
The modern obsession with originality may be the exception in history rather than the rule.
In Jefferson’s day, the concept of authorship was often looser than it is today. (Jefferson drew heavily from George Mason in writing the Declaration.) Writers borrowed freely from classical sources, sermons, pamphlets, letters, and one another. Footnotes were rare. Formal citation standards had not yet become the norm. Readers cared less about tracing every idea to its origin and more about whether an author had faithfully assembled existing wisdom into a persuasive and useful argument.
The Declaration succeeded not because it was original, but because it distilled and organized ideas that many Americans already believed.
The eighteenth century had a different understanding of authorship. Jefferson’s greatness lay not in inventing every idea he expressed, but in assembling, refining, and articulating them. AI-assisted writing may be pushing us back toward a similar understanding of authorship—one that emphasizes judgment and synthesis over strict originality.
Every generation writes with the tools available to it.
I often wonder what Thomas Jefferson would have thought of ChatGPT. A man who spent a lifetime collecting books and organizing knowledge would likely have been fascinated by a machine capable of conversing with a library.
Jefferson had a quill and a library.
I have a keyboard and an LLM.
The responsibility for what appears on the page remains exactly where it has always belonged—with the author.
AI Disclosure: Thomas Jefferson had a quill and a library. I have a keyboard and an LLM. This essay was developed with the assistance of ChatGPT as a research, editing, and drafting tool. The ideas, conclusions, and responsibility for the final work remain mine.
