On March 31st, I decided to post every day on my newsletter ADHD Success with AI.
What does that mean?
I created 5-ish weeks of posts, one per day. A total of 33 posts.
I scheduled them to submit every day at 7 AM.
How much content did I have queued up?
None
How many post ideas did I have lined up?
A handful, I didn’t touch them.
Why would I even think this was a good idea?
Because it seemed like a fun experiment.
It all started by questioning the value of newsletters
I’ve had conversations with several friends recently about how much they read, or fail to read. Or about how many newsletter subscriptions they have. Or why they subscribe to so many.
I don’t know anyone with more than 3 newsletters who are reading all of them.
Substack is trying to be all things to all people, so if you’re here, you likely feel compelled to subscribe to a newsletter, or all of them if they say something clever in notes.
But once you’re here, you see ‘reels’ or videos or live streams. You see Notes and you see podcasts. You can do just about anything here, and its hard to say which thing is right for you.
More importantly, once you’re here for a while, you start to see patterns and posts and topics repeating themselves. You also end up in a niche by accidentally following or subscribing to certain posts.
It all came to a head when I came across a newsletter that was new to me, that had over 1k subscribers, but no “name” as an author. Then I opened up a newsletter and saw that it was very clearly written by AI, maybe with some minor editing.
I’ll say it again…
I found a 1k+ subscriber newsletter that was clearly written with ChatGPT.
Long story short, there are lessons and thoughts…
I started this at the start of April, it’s now the 15th. I learned a LOT at the start, and have some mid-month insights as well.
Being “authentic” is NOT going to happen with an email written by a robot. Sharing information, that can still be useful, as long as it’s still accurate and/or helpful.
Something created by a robot hits different for readers, but also hits different for a “writer.”
ChatGPT doesn’t complain, it doesn’t roll its eyes, and it doesn’t sigh when you ask it to improve something.
It seems easy, but it’s actually front-loaded work. I spent around 10 hours creating 33 posts, setting up automatic posting, and adding images.
Just posting without “marketing” isn’t enough. Substack isn’t advertising your posts, and it isn’t pushing them on anyone. If you’re starting from 0, don’t expect rapid gains.
You CAN make an automatically generated newsletter based on YOUR content, and in a voice that is very close to your own.
I’ll be detailing what I did closer to the end of the month. I think this isn’t as sketchy as it sounds, and have my own ‘ethical take’ on the process.
Everybody is on their way… “Winter is Coming”
The economy, globally, is headed in a dangerous direction. Not just because of the “tariffic king of america”, but also because AI is teaching itself to be better.
By some estimates, we’re 2–3 years away from a thinking AI (AGI.) By this time, almost all “code” written will be done by AI.
At this point, an AI could do just about any white collar work that we’ll give it (eat the c-suite!)
In around 6 we’ll have AI smarter than the cumulative intelligence of all humans.
We will become, if we’re lucky, stewards of that intelligence. Using OUR creativity, assisted by AI to pursue every idea under the sun.
I say all that to say, jobs will be disappearing, people will be trying to get rich, or just survive, by leveraging AI for help.
I won’t have more attention to read more newsletters, I’ll be getting very picky in what I read. If newsletters can start to be a by-product of interesting thoughts, even for non-writers, I’m here for it.
So, I’m doing this to be better at AI, and to teach people to be better at AI.
Follow along here and I’ll report back later.
Go read that newsletter and see how “AI” it feels. Complain about it, marvel at it, but mostly think about what it means for the future of newsletters and writing and general help.
Great experiment Jody! I can’t wait to see what you learn in another 15 days. Just reflecting on this is shifting my thoughts. Now if you could build me a GPT to read and summarize all the newsletters I’m subscribed to…
Then my AI could read your AI 🤣
I've been toying around with creating a "Document Management" AI-generated substack newsletter -- idle curiousity from my past life. I think a couple three things here:
1. Very cool experiment
2. Folks continue to wildly over-estimate how quickly AI is going to be implemented. And from my 30 years IT-adjacent, the "70% implementation failure rate" of IT projects keeps popping into my head. There's just not enough talent to go around to do this goodly, much less quickly and goodly.
3. I think AI is a perfect assistant for curation type newsletters, but that's simply an elevation of newsfeeds, not a new use case.