Can you love creativity and still use AI?
Exploring the messy, complex tension between discovery, creativity, and the ethical use of AI.
Recently, several writers I admire — real writers, with decades of experience, their creative hearts poured into every line — discovered that their published works had been scraped into a massive unauthorized database, LibGen, to train AI models.
If you’re a regular Substack reader, you’ve probably also come across these authors’ stories of horror and anger at the discovery of their work being stolen. Maybe you are one of these authors.
As I read the various posts and their comments, I found myself nodding along, but also feeling challenged. Each post I read made a very compelling argument that training AI models without their consent was nothing less than theft. They expressed their concern for the loss of human creativity, the writer’s voice, and a copy-and-paste culture developing in the writing community.
I agree, I share their concerns, and I think their anger is absolutely justified.
And yet.
Here I am writing about the value of AI in education and lifelong learning in my AI Diaries Series on this Substack. Here I am, using AI to help clarify my thinking, to find weird and wonderful facts about marine iguanas for my son’s show-and-tell, and to even help me write these posts.
How could I click the 💛 on these authors’ posts without feeling like a hypocrite? How could I comment, knowing I might be seen as aligning with the very forces that harmed them? Was it okay for me to be using AI in any form if I truly valued human creativity?
That tension is why I’m writing this today.
To be transparent.
To be honest about my own use of AI — and how that sits alongside my creativity, my work as an educator, and my fervent desire to be a good human.
The muddy waters of AI
I think part of the tension arises from the murkiness of the question, “What is AI?”
In today’s world, “AI” can mean anything from image generators to podcast audio cleanup. Even putting a green screen behind your Zoom background now counts as using "AI."
Part of that is just marketing hype from software companies jumping on the AI bandwagon, but this muddy labeling makes the conversation harder.
It’s no longer just about ChatGPT writing essays. It's about everything from accessibility tools to creative theft — and everything in between.
This debate can’t be summarized into pro-AI and anti-AI. This is nuanced stuff.
My early skepticism
When ChatGPT first appeared on the scene in late 2022, I was verrrrry skeptical.
I worried about where its training data came from. I worried about inaccuracies and what I now know to be hallucinations.
As a copywriter, I was a bit worried that it might threaten my livelihood, but I wrote it off as a fad. I figured we needn’t bother about it, because it would soon go out of fashion. *taps foot…still waiting for that*
But then, I attended a workshop on ChatGPT in education by Brandon Hendrickson.
And I think that’s the first time I realized the possibilities of AI.
Sure, you could use it to cheat and shortcut. But it’s way more interesting and useful if you use it in an intentional way.
After that workshop, I discovered that AI offered us new ways of presenting information, of uncovering the wonder of the world beyond the screen, of stirring up curiosity and creativity that is profoundly human.1
I started to wonder: Could AI be another tool in our learning toolkit? Could it guide us to new sources — books, documentaries, academic papers — that would deepen our knowledge of a topic?
Was it possible that AI might not be a substitute for creativity, but instead be a catalyst for wonder?
Acknowledging the real harm
I don’t want to gloss over the real harm being done by unethical AI development.
When companies like Meta scrape copyrighted works without consent, it’s not a small issue. It’s theft. It’s exploitation. And I stand in complete solidarity with the writers who are furious — they have every right to be.
I also believe that ethical concerns should not be used as an excuse to abandon the conversation entirely. If we, the ones who care about creativity, discovery, and ethics, refuse to engage with AI at all, we leave its future development to people who may not share those values.
We don’t make the world better by vacating the room.
Instead, I’d like to see more creative, curious, good people using, shaping, and benefitting from AI.
Why I believe in good people using AI
In my work as a copywriter, I’ve encountered so many people with brilliant ideas and sought-after skills who don’t see themselves as “writers.” AI can help them craft and clarify their messages to make their voices heard. To me, this is an example of AI supporting creativity and amplifying voices we wouldn’t have heard before.2
As a homeschool parent and mentor, I’ve seen how the intentional use of AI can help kids see the whole world as interesting and worth caring about. I’ve seen barriers to learning being broken, and the eye-sparkles and giggles as my kids discover something cool. As an educator, this is nothing short of delightful.
And then, as a multi-passionate, multi-hyphenate, autodidactic, homeschooling mompreneur3… AI tools help me to get started, clarify my random thoughts, and hit that publish button — connecting me to you.
AI holds so many possibilities for creativity, curiosity, connection…
I don’t want the unethical practices of big tech companies to rob us of all of these amazing possibilities. I don’t want AI to become a tool only for those who seek to exploit or profit at others’ expense.
I want it to be a tool for amplifying goodness, creativity, and curiosity.
Driven by discovery
This Substack is called Discovery Diaries for a reason.
I am driven by curiosity, and I believe that discovery and learning are lifelong processes. I’m on a mission to fall more deeply in love with the many wonders of the world, and I want my kids to feel the same way.
Used carefully, AI has become one more tool in accomplishing that mission.
A conversation partner.
A clarifier.
A co-explorer that could point me toward books to read, ideas to research, and questions to ask — but could never replace the slow, joyful process of learning and creating.
(And for the record: when you ask ChatGPT to tell a joke, it's terrible. Human creativity always wins when it comes to humor — even the Dad jokes.)
How I actually use AI
So now, you may be wondering how I use AI in my writing and my creative process more broadly. It varies, and is always evolving, but here’s my typical process:
1. Speak it out and get a transcript
I start by speaking out my ideas into a tool like Descript4. Sometimes I’ve got something clear to say, but more often I just let myself talk through an idea.
I get a transcript that’s full of tangents and half-thoughts. But invariably I’ve also got a few gems in there too — these tend to emerge around two-thirds of the way in when I’m hitting my stride and feel like I want to give myself an, “Amen Sister!”
2. Get ChatGPT to tidy and structure the transcript
I bring that transcript into ChatGPT and give it strict instructions: Correct transcription errors, remove vocal ticks, remove blatant redundancy, and add subheadings to bring some structure to the ramble. But, DO NOT write anything new.
The output is a shortened, structured piece of writing that still retains my voice (it originated with my spoken words).
3. Develop the first draft
At this point, I might take the output and start shaping and editing it into a post on my own.
Or, I’ll get ChatGPT to create a first draft of a post for me. In this case, I give it detailed prompting and I tell it to prioritize my turns of phrase, stories, and examples.
Then, I go into full editor mode. Actually, I don’t think the term “editor mode” is even correct — it’s more like senior writer mode — because I do a lot of rewriting and restructuring.5 I think if I was editing an actual writer’s work, I would be fired for doing too many rewrites and being infuriatingly pedantic!
4. Edit — for real this time
Along the way, I use tools like Grammarly to check spelling and grammar. And if I’m not completely exhausted, I’ll run the final piece through ChatGPT, instructing it to only list any blatant errors.
5. Heading and subheading
Once the main body of the post feels solid, the next big hurdle is capturing attention with a strong heading and subheading.
Heading and subheading combinations are where I unashamedly put ChatGPT to work.
Here’s the thing: the heading is so important — it’s what’s going to make you open your email and read this post, or — if it’s really good — make someone stop the scroll and click on my article on Substack Notes.
Headings are SO DIFFICULT to nail. However, I’ve found that quantity helps. And ChatGPT can do quantity.
So, I’ll use the same chat window I’ve been brainstorming in, and give it some context about the angle and the problems my article might solve for my readers — and then let ChatGPT have at it. I start by asking for five options, but then often ask for five more.
And then, because I cannot be told, I’ll usually go with a hybrid of a couple of the options.6
→ AI is a helpful tool
Hand on heart, I can assure you that the soul of my posts — the story, the experience, the spark — is always mine (or my husband Garren’s, who is along for this family home education journey).
AI just steps in as my clarity partner.
It’s the tool that helps me move from a passionate but rambly notion to a still-passionate, coherent story. And it helps me confidently hit publish, rather than letting my writing sit languishing in a drafts folder.
I like to think that the improved coherence and clarity — combined with my real human experiences — is in service to you, my reader.
My commitment
So there we go, I’ve said what I’ve needed to say (for now at least — this is a continuing conversation).
For now, I’m (re)committing to:
Intentional use of AI, and being transparent about when and how I use AI.
Continuing to advocate for ethical AI practices.
Human creativity, curiosity, and a love of learning and new discoveries — with AI being a support to these priorities.
Being guided by my mission to help more people — especially kids — fall in love with the wonders of the world.
An invitation
This isn’t a finished conversation.
I certainly don't have all the answers.
But I’m committed to asking the questions (even if they are uncomfortable), and to finding a way forward that honors both creativity and the possibilities AI tools offer us.
If you have thoughts, tensions, stories — I would love to hear them in the comments.
We’re trailblazers in this new world. Even if we have differing views, I think it’ll be easier if we work together.
Because the future isn’t just about technology. It’s about what kind of humans we choose to be.
Brandon will give most of the credit for this to the educational philosopher, Kieran Egan. And of course, I’d agree with him. But we’ll save that story for another post. In the meantime, you might be interested in checking out Brandon’s “Egan and ChatGPT Walk into a Bar” series.
This was partly the inspiration behind my course, Words+Web+Wisdom, where I coached a group of business owners to write their own website copy, using ChatGPT — without sounding like a robot.
Phew, that was a mouthful of frankenwords!
Descript is a very cool tool with plenty of “AI” features! I usually use it to edit podcasts, but in this case, I’m just using it as a transcription tool.
With all the refinements I do, I sometimes wonder whether AI saves me any time at all! But I do think it makes my writing better, and I do think it helps me to start and finish my writing more than I did in the past. Either way, I don’t agree that AI is making me a lazy writer.
I’ve started developing a list of AI-clichés I’ll chuck out immediately — transform and unlock are two that jump to mind. I’m also noticing it’s using “spoiler…” a little too liberally lately!