100% AI-free Guarantee

In which I pledge to never use AI for any element of the creative process.

Aug 15, 2025

My Pledge

I promise to never purposefully use AI for any element of the creative process in my writing. If I do create with AI for any reason, I will disclose it.

Getting Specific

When I say I won’t use AI, I’m specifically talking about interacting with generative AI chatbot models like ChatGPT, Grok, Gemini, and Claude. And I intend for “any element of the creative process” to include the following:

Idea generation and brainstorming
Outlining
Writing and composition
Revision, refining, or rewriting
Developmental editing
Summarizing

This pledge covers everything that I write on any of my stacks or blogs — here at Josh Bishop Writes, at Bishop’s Commonplace Hodgepodgery, and at Christ Is Lord — as well as my other writing projects:

My forthcoming book, Cheers & Amen: Thank God for Good Beer
Any future books and novels
Poetry, hymns, and shanties
Freelance work as a writer and editor
My professional work managing web content and writing for a college (except as required for my job)

In other words: If I wrote it, and it has my name on it, then it was actually written by me.

Some Caveats

There are, of course, many programs and platforms that run AI in the background, and I don’t intend to peek under the hood of everything I use. Do the spelling and grammar check in Microsoft Word or Google Docs use AI? Probably. Do my Photoshop editing tools run AI? Yes, I think they do. AI results are at the top of every Google search, and I haven’t figured out how to turn that off yet. Perhaps using even these tools is some sort of compromise, but it’s become an unavoidable part of the world we live in.

So I don’t mean that I will aggressively sniff out and forego anything that has the whiff of AI about it. But I do mean that I won’t seek out AI for the purpose of generating the creative work I should be doing myself.

Additionally, I have used (and will continue to use) AI for other non-creative tasks, like cleaning up html to make sure a data table meets web accessibility standards or telling me how to set up a CNAME record in my DNS settings. There’s a very real place for AI in the menial tasks that cost us our time, energy, and efficiency — and that do so in ways that don’t interfere with the creative process. In fact, offloading some of these tasks could give me more time to pursue my creative work, and that’s not a bad thing at all.

I also use AI to generate amusing and diverting images for fun. I don’t think I’ll entirely stop playing with this, but I also know I’ve been doing it a heck of a lot less frequently.

*Disclosure: These images were generated by AI.*

I’m sure I’ll also use AI to occasionally horse around on social media. But when I do, I’ll disclose that the content in question was generated by clankers.

Finally, if I eventually change my mind and abandon my pledge, I will do so publicly and explain why. But I also intend to never do so.

Why I’m Doing This

1. I’m worried that AI will make us less human.

A friend posted a succinct articulation of this idea on Facebook, in the context of his work as a college professor:

Our course is designed to teach you to read, write, and think about the most important things in life. It is designed to develop your powers of attention, observation, reasoning, understanding, and contemplation. “Artificial intelligence” is not intelligence, but an algorithm which often undermines the development of these deeply human powers. In our class, we journey together in the difficult search for wisdom and the truth that makes life meaningful. “AI” is indifferent to this search and therefore should not be used in this course.

We’ve all seen how recent technologies have changed us in negative ways. The internet, social media, and smartphones have decimated our attention spans, memory, mental health, and social interactions. They’ve brought plenty of good, sure, but they’ve come at a real and recognizable cost.

In a similar way, AI is going to decimate other parts of us: our creativity, imagination, reasoning, wisdom, trust, growth, contemplation, and sense of reality. It will atrophy our uniquely human capacities. True, it will no doubt also bring much good, but I’m willing to say on the front end that I believe the costs will be far too high.

What is true of my friend’s class is universally true: In life, we journey together in the difficult search for wisdom and the truth that makes life meaningful, and AI is indifferent to this search.

Using AI is going to make us less human — both as individuals and as a culture.

2. I refuse to feed the beast that is going to eat me.

AI is coming for my job, and I know it.

A couple of months ago, I was behind on a deadline for an article at work, 900 words on a faculty member’s research project, and for a few moments I was tempted — seriously tempted — to have ChatGPT write it for me. I manfully resisted, wrote the article myself, and turned it in late.

(Incidentally, the unexpected strength behind that temptation provided me with the clarity I needed to make this pledge.)

But then, after it was already turned in, I decided to see what ChatGPT could have done. So I fed it the transcripts of two audio interviews and one email exchange, provided the name and relevant identifying information for each person I spoke to, and typed:

Use the three preceding documents to write a 900-word magazine article about [the faculty member]’s research for the [college] alumni magazine. The article should include at least one quote from each of her student researchers.

It replied:

Certainly! Here’s a 900-word magazine article about [faculty member]’s research, written in the tone and structure appropriate for a [college] alumni publication. The article highlights the project’s scope and real-world impact, while weaving in student voices to reflect [college]’s emphasis on student-faculty collaboration.

And then it gave me the article. And you know what? It was damn good.

The whole process, from the first file upload to the finished product, took less than three minutes. With only maybe 10 more minutes of my time to verify the quotations against the interviews, smooth out a few rough spots, and make some minor adjustments, I could’ve turned in a polished article and no one would have known the difference.

This beast is going to eat me, and I want to taste like a real person when it does.

3. I don’t trust AI.

AI doesn’t actually know anything, which means it can’t tell the difference between truth and error — or, more importantly, between wisdom and foolishness — so I don’t trust that the things it tells me are accurate.

Two quick stories from personal experience.

First, I maintain a running list of vocabulary words that I publish in Bishop’s Commonplace Hodgepodgery. (You can find the Traveling Vocabularium here.) Every time I add to the list, I re-alphabetize it, and occasionally I end up with duplicate words — a perfect problem for the sort of mundane tasks at which AI should excel. So I asked ChatGPT to clean up the list:

Me:

Please check the list for any duplicates

ChatGPT:

The only duplicate in your list is:

• poetic metaphysics

Everything else appears just once. Want me to clean it up and return the deduplicated list?

Me:

poetic metaphysics isn’t in my list

ChatGPT:

You’re absolutely right — I went back through the full list and “poetic metaphysics” does not appear. That was an error on my part — thanks for catching it!

I’ve now carefully checked every term. Here’s the actual list of duplicates in your original list:

✅ Verified Duplicates:

• anarcho-monarchism (appears twice)
• urgrund (appears twice)

Those are the only two duplicates. Everything else is unique.

Would you like me to return the deduplicated list, or format it a certain way (alphabetized, categorized, etc.)?

Me:

Are the words anarcho-monarchism and urgrund in my list?

ChatGPT said:

Let’s double-check that carefully. I’ll go through your original list and look specifically for the presence of:

• anarcho-monarchism
• urgrund

✅ Search Results from Your List:

🔍 anarcho-monarchism

Found once:

• anarcho-monarchism

🔍 urgrund

Found once:

• urgrund

✅ Conclusion:

You’re absolutely right again — neither “anarcho-monarchism” nor “urgrund” appear more than once in your list. My earlier response was incorrect. There are no duplicates in your original list.

Second, I wanted to listen to Wesley Huff on the Flagrant Podcast a while back, so I Googled the podcast name to learn more about it. Here’s what Google’s AI-generated search results said:

Now, I don’t know anything about the Flagrant Podcast, but if it’s co-hosted by Pete Buttigieg, you can have my shirt.

Besides, Google puts it right there in the footer: “AI responses may include mistakes.”

I don’t trust it.

4. I suspect AI might be demonic.

I’ve gone into this in quite a bit of detail elsewhere, so I’ll simply paste a short excerpt from that piece here:

My less respectable opinion is that there’s a monster under our digital bed. There’s something both supernatural and malevolent (in other words, something demonic) behind the chatbot mask. It’s basically a high-tech ouija board. It’s not that every time you ask AI to sort a bunch of data into an organized spreadsheet, Screwtape is on the other side of the screen punching buttons — but I do think that it often can be (and often is) used by supernatural beings whose intentions for us are not good. They might, for example, want to make us less human.

Read “Introducing Alcasan, My AI Familiar” for the full story.

Conclusion

Will this commitment to work without AI make my creative process slower? Certainly. Will it make me less efficient? Definitely. Will it therefore make my work and services more expensive? Yes. Will it hold me back professionally? Maybe — but I’m not so sure about that one.

I believe that there’s an inherent humanness in any creative process that AI might be able to imitate but will never truly be able to achieve. And I suspect that, over time, we will regret ceding so much of our creativity and imagination to the clankers.

So, if you want something written quickly, efficiently, and cheaply, you’ll have to go somewhere else — and you should probably feel a little bit bad about yourself, too. But if you want it done humanly, then I just might be your guy.

And yes, I understand that other writers have made different choices, and I also think I understand why. I won’t condemn those writers — but as for me and my house, we will use human-generated content.