Resistance in Writing

And where AI fits into all of this

The Voice That Tells Me Not to Start

Steven Pressfield calls it “Resistance”—the invisible force that shows up every time you try to do meaningful work. It’s the voice that says, “Don’t start yet. Wait until you have a better plan. Someone’s going to think you’re dumb. Someone else has already done it better.”

Lately, that Resistance has been palpable when I think about writing.

What if I say something that gets misunderstood? What if someone with a big platform thinks I’m stupid, quote-tweets me, and everyone piles on? What if I say something true about my life or faith and someone I respect sees it and just… silently judges me?

It really only shows up when I want to publish something (i.e. I feel no Resistance to journaling), and the antidote seems to be just getting my butt in the chair.

The Friction That Keeps Me From Publishing

There’s also a logistical side to Resistance.

Where do I publish? Twitter? Substack? LinkedIn? Should I use my real name or stay anonymous? If I write about faith, how personal should I get? If I cross-post a Substack article on Twitter, will it get buried by the algorithm? Will people think it’s annoying if I post it once and then retweet it later to combat the nerfing? What’s the best time to publish?

None of this is crippling on its own. But the decision fatigue builds. Each question adds a bit of hesitation. It creates just enough friction to keep the draft in the Word doc—or stuck in the GPT thread.

Writing as Processing, Not Just Publishing

I actually love writing.

Writing helps me clarify what I believe, and what I feel. It helps me move from mental noise to something more grounded. Sometimes I do that by journaling by hand. Sometimes I type in a Word doc. And increasingly, I riff out loud into GPT using the voice feature. It helps me get unstuck, get the raw material out, and then shape it into something I can share.

That act—just saying the thing—makes the fear smaller. But there’s still a quieter Resistance that shows up later, when it’s time to think about sharing what I have written. This gets amplified when AI is involved.

A New Kind of Resistance: When AI Makes It Easier

GPT has been incredibly helpful for me. I can riff a raw idea out loud, and GPT helps clean up the grammar and structure. The idea’s still mine. The voice is still mine. I read and edit before posting. But somehow, I still feel a weird sense of guilt—like it didn’t “count” because I didn’t suffer enough for it.

Samir Patel from Askeladden Capital recently wrote about how AI is going to drastically accelerate his research process, driving 2-3x productivity gains. And I believe it. We all know these tools can help us think faster and ship more. But I think a lot of us are afraid of actually leaning in. Not because they don’t work—but because they do. Because stepping into that level of leverage forces us to let go of old habits, old excuses, and old versions of what “real” work looks like. There’s an element of letting go of control, too.

There’s something oddly vulnerable about letting a tool help you. I think it requires a certain humility to acknowledge that something anyone can access—at little to no cost—can now do significant chunks of your job better than you. It’s like asking a taxi driver what he thought about Uber back in 2015.

So what does a writer’s role become in a world where AI is fully integrated? I’m not entirely sure. But if I had to guess—it’s less about polishing every sentence by hand, and more about thinking clearly, generating good ideas, and sharing them more consistently. It might mean publishing more prolifically, using the tool to raise the baseline quality of each piece through faster iteration, and spending more time on the part that really matters: the thinking. In that light, AI isn’t a threat—it’s leverage.

Did I Still Write It?

Sometimes I ask myself: if I speak the ideas and GPT shapes the sentences, is it still fair to call it my writing?

I think it does. If the spirit and intention are mine—if the ideas are real, and I stand by every sentence—I think that’s authentic. The process may look different, but the ownership is still there. And if it helps me share something meaningful, then it’s not just useful—it’s good.

That said, I don’t think I need to slap a disclaimer on every post. But I do want to be transparent about the process.

Write Anyway

So in spite of all the hesitation and friction in everything above—here’s where I keep landing: I still need to write. I need to hit publish.

Doing anything in the face of Resistance is an act of faith. Even if a small one. It’s an action that proclaims that God is bigger than fear. Choosing not to publish is an action that proclaims that the failure and criticism I fear is the authority I’m ultimately trusting.

So I write. In journals. In Word docs. Into GPT. Sometimes I polish it all myself. Sometimes the model helps me shape it. But in all cases, the thoughts are mine. The tension is mine.

The goal isn’t perfection. It’s honesty. Connection. Clarity.

Appendix: A Few Additional Notes

a) I considered including the ChatGPT thread where I wrote this but it’s 15k words and I’d have to edit out some personal info. If that would be interesting to see let me know.

b) I wonder if the first iteration GPT produced would have been an equal or better experience for the reader. And whether my incremental edits add any real value. It does make it feel more authentic. But, substantively, I don’t think the message changed. It’s humbling.

c) I am still not sure of the best way to use AI to help me write. The way to work with AI in writing changes depending on what you’re writing. An equity research report is probably going to follow a different approach than a blog post. It’s a bit harder to get it to pull from disparate sources correctly, proportionately, and with a good sense of what is important, as compared to just cleaning up a riff and suggesting better ways to phrase things.