On AI, Blogging in 2026, and Other Creative Conundrums

So, the big conversation topic in the blogosphere this week so far seems to be the new AI detection feature Substack just implemented. It's powered by a company called Pangram, and it's causing quite the uproar for all the reasons anyone already familiar with this type of tool might expect:

• False positives (and to a marginally lesser degree, false negatives)
• Worries about their content being used to train Pangram's AI without permission
• People continuing to lie their asses off about their often obvious AI use anyway

Although I'm not even sure why we're still talking about AI detectors at all in 2026, because truth be told? AI detectors are complete scams and absolutely full of shit, and I honestly thought everyone knew this by now.

We freelance copywriters found that out a couple of years ago

AI writing tools like ChatGPT and Jasper were first catching on at the time. And we remember all too well how everyone was suddenly concerned about the possibility of human writers handing in AI-generated content instead of actually doing the work they were paid to do.

Clients started accusing even writers they'd been working with a long time of using AI to complete assignments, freelance writing platforms started applying mandatory AI content scans to everything members turned in, and all that good stuff. You found out real quick just how easy it was for even purely human-written content to royally flunk a detection scan.

You also learned real quick how to make sure your content passes, whether it actually deserves to pass or not.

Just saying.

I get all this is supposed to encourage transparency

It won't, though. The existence of these things does nothing to deter people from lying about AI use.

Literally every single day, I see countless posts on social media, LinkedIn, Medium, Substack, and just about everywhere else that read like textbook AI output. Yet if confronted, the person swears up and down that they don't even touch AI, let alone write with it (because social consequences).

Even the stupid ones know that AI detectors are full of shit, so they also know all they really have to do is deny that they use AI.

And the smarter ones who want to make damn sure they don't get accused of AI use in the first place? They not only understand how to make sure their content tricks AI detectors into thinking it's human-written, but they also know how to make sure their writing reads as human enough that you won't even question it.

Substack gets all this.

But it's doing what every company is doing these days – its very best to make sure it looks like it values authenticity, wants to ensure quality, and cares about the same things its users care about. And maybe it does. But I can assure you that Substack's team cares a lot more about preserving its reputation as the one long-form writing platform people can still go to for quality indie writing, whether that's really true or not.

My own stance on AI use

The one thing that actually struck me as potentially useful about this latest round of Substack changes was the option to add an AI transparency statement to your publications (which I may actually do eventually).

I don't necessarily feel I need a statement or anything, as I've always been extremely transparent about AI use, as well as what I do and don't use it for. But in 2026, we're probably traveling into territory where posting one would be a time-saver, as far as briefing anyone new.

Because while I completely understand why someone might want nothing to do with AI (and respect their right to avoid creators who don't feel the same), I'm not an anti-AI creator. I find AI fascinating, and I've written about it (professionally and otherwise). I also generally advise people to educate themselves on how it works and at least learn how to use it, even if they ultimately choose not to.

That said, writing isn't just what I do. It's a huge part of who I am as a person, and I enjoy the act of writing too much to outsource that part of creating to AI. But I do use various AI tools to:

• Brainstorm new ideas, especially if I have writer's block
• Develop half-baked ideas I already have well enough to move forward with them
• Create content outlines and summaries to work from, especially for long-form content
• Create or refine images and illustrations to accompany posts
• Unpack why a given post might have performed better or worse than average
• Research or otherwise gather information for possible inclusion in a post

The only time I leave the job of actually writing something to ChatGPT or a similar tool is when we're talking about fringe stuff I don't want to waste my creative energy on – like metadata, basic product descriptions for my shops, or some of the blurbs that accompany long-form content if I share it across social media.

I don't actually edit or refine my content with AI, though.

I'm anal about the way my stuff reads. I swear or sometimes go off on minor tangents, and AI seems bent on sanitizing anything that sounds the least little bit off-color. So, yeah, I want the final say on how my finished writing sounds.

Probably always will.

I don't really know how bad or good Substack's detectors are yet

I have yet to compose anything new since they implemented it a couple of days ago, so I can't really speak from personal experience. But I've read enough posts from fellow Substackers to feel relatively confident that this detector will be as useless as every other I've encountered so far.

So my advice on navigating stuff like this moving forward is basically this.

If you're a writer

When asked how I feel about the future of writing, visual art, and similar disciplines now that AI has entered the chat, I'm generally in favor of people at least learning a couple of the popular tools (whether they decide to use them or not).

I'm old enough to have seen a lot of tech trends come and go. And although I do think we'll see a lot of changes in how we regulate the technology moving forward (and rightly so), AI isn't going anywhere. It saves big, powerful, important companies way too much money and time for it to go anywhere, so the rest of us need to learn to live with it.

The smart thing to do would be to figure out how to use it to your advantage. What that looks like will vary a lot from one person to the next, but yeah.

At the very least, AI can be super helpful for research and organization purposes. But be transparent about how and why you use it. I swear the internet is getting faker by the day. That makes honesty, authenticity, and all the little flaws that make us human valuable, arguably for the first time since... like... ever.

Use that.

And if you're a reader

It's probably in your best interests to stop splitting hairs as far as any given creator's AI use, because mark my words. If someone really doesn't want you knowing they used AI to create something, you won't.

I dislike reading other people's lazy writing as much as the next guy, and purely AI-generated content with zero human refinement absolutely reads as lazy writing. It's full of circle talk and that super irritating cadence ChatGPT and similar tools always fall back on.

It bores me at best, just like really shitty human-created writing does. But I've also read a lot of AI-assisted hybrid content that's really pretty informative and otherwise everything I like my reading material to be.

So I focus mostly on two things:

• Is the content actually any good? Is it doing a good job of entertaining me, informing me, inspiring me, or just plain holding my attention?
• Is the creator transparent? Am I picking up on AI use they're lying about?

Because we're moving toward a world where the content we consume is increasingly hybrid in nature. This is happening so swiftly already that conversations like this one won't even be relevant anymore in a few more years.

I'm doing what I can with that.

Eventually, everyone else will have no choice but to do the same, like it or not.