Looking for advice in all the wrong places
ChatGPT helped me figure out what I knew all along.
Over the past few months, I’d been struggling to keep up the pace of this newsletter. I couldn’t figure out how to write a new essay each week, respond to reader emails and comments in a timely way, and consistently find new stories and inspiration to inform newsletter topics — while also freelancing full-time and raising a toddler. I somehow managed to produce something each week, but I could never shake the feeling that it wasn’t good enough. That if I had a little more time, my writing would be better, my takeaways stronger.
This newsletter holds a permanent spot in my busy mind. I’ve debated how much of myself to share, how much of my time to devote, how much effort was too much. I’ve considered all the balls I was juggling and which ones I was inevitably dropping. I’ve wondered if it would be possible to once again view My Sweet Dumb Brain as a source of joy, not apprehension.
These thoughts required deep consideration. But, as much as I tried, I couldn’t figure it out in real time, alongside all of you. So I stopped writing. In mid-May, I took an unplanned newsletter pause and looked to the universe for answers.
The universe, as it turns out, didn’t have any answers for me — or, at least, not as quickly as I wanted them. In my first week away from My Sweet Dumb Brain, there were no obvious signs, notable epiphanies, or unexpected opportunities to give me the direction I needed. So, I went to ChatGPT instead.
That’s right. In a moment of existential and creative struggle, I looked to artificial intelligence for guidance.
For the uninitiated, ChatGPT is a text-generating AI chatbot. When given text prompts, it’s able to write essays, press releases, novels, code and more. ChatGPT and other AI technologies, with their ability to outperform humans in specific activities, are a source of worldwide controversy and endless ethical debate. They’re also, as I discovered, pretty fun to experiment with.
Things started out innocuous enough. I’d been wondering if I’d have more time and energy for this newsletter if I did things like preparing meals ahead of each week or making better use of my mornings. Optimizing my life, if you will.
Before taking a family trip to the grocery store, I pulled up ChatGPT and asked for quick lunch ideas that would yield leftovers. It suggested a quinoa salad “with fresh vegetables like cucumbers, cherry tomatoes, bell peppers, and red onions.” We bought exactly those things. My partner added feta to the mix, along with ChatGPT’s suggestion of fresh parsley and a simple vinaigrette.
The salad fit the bill. It was a little bland, but gave us plenty of leftovers. I ate it for lunch a handful of times that week. Easy peasy.
The next day, I decided to ask AI to provide a calm morning schedule that would allow me time to get ready, squeeze in some work, and make coffee — all ahead of my daughter’s usual wake-up time. “Of course! I’d be happy to help you plan a calm and refreshed morning routine,” the bot cheerily replied.
ChatGPT’s proposed schedule had me getting up at 6:45 a.m., an hour earlier than the usual time I’d get up with my toddler. “Give yourself a few minutes to wake up and stretch,” it instructed me. “Take a deep breath and set an intention to have a calm and productive morning.”
The rest of the schedule didn’t exactly make sense — ChatGPT suggested I eat breakfast with my daughter at 7:30 (“This is a great time to bond and have meaningful conversations,”) but to also get ready for the day at 7:35 while she was eating. Still, it gave me the structure I needed. I made some adjustments to the itinerary and got out of bed as planned at 6:45 the next day. By the time my toddler woke up, I’d already made a cup of coffee for me, a bowl of oatmeal for her, and even squeezed in a half hour of work. Calm and productive, indeed!
By the middle of the week, I was feeling good about the specific prompts I was giving ChatGPT and the direction it was offering me in return. On Wednesday, armed with AI-generated questions that would lead to a meaningful discussion, I took my mom out to breakfast as a belated Mother’s Day gift. My mom and I don’t typically struggle for conversation topics, especially during one-on-one time, but I figured the prompt wouldn’t hurt.
Out of all of ChatGPT’s suggestions (“How has retirement been treating you?” “Is there a particular book that has had a significant impact on you?” “What was your favorite childhood memory?”), there was one I especially liked and asked: How has being a grandmother changed your perspective on life?
My mom’s answer was so lovely and heartfelt that I experienced an immediate pang of guilt knowing that the inquiry wasn't my own. “Oh!” she said, her face lighting up as soon as I asked the question. She talked about how seeing the world through her granddaughter’s eyes has made her more hopeful. She admitted that she thinks about how different things would be if she didn’t have such a joyful, young perspective to balance out her cynicism. She talked about her dreams for the future. As she spoke, she seemed lighter — like she’d been waiting all along to talk about just this topic.
When she finished, I sheepishly told her about my experiment: that I’d been looking for guidance from an artificial intelligence bot. She seemed a mix of confused and surprised. We both agreed it was still a great question.
As the days went on, my success with ChatGPT was limited. A work-from-home schedule that incorporated exercise was almost identical to the schedule I’d already laid out for myself. Ideas for preparing my toddler for summer camp were laughably off the mark. Suggestions for interesting podcast episodes to listen to were outdated and generic. It didn’t take long before it felt like I was wasting time trying to come up with the perfect prompt that would generate the answer I was looking for.
By week three of my newsletter break, I’d given up on ChatGPT entirely. Instead of spending my time tweaking the prompts I was creating, I started going on longer, more frequent walks — a tried and true way to generate ideas. It was on those walks that I finally found the epiphanies I needed.
I don’t want to stop writing My Sweet Dumb Brain. I love the exercise of writing, the feedback from readers, and the fact that so many of you take time out of your busy lives to read my words. But I also can’t keep up the pace of producing an essay like this one each week — and no matter how impressive ChatGPT might be, it’s not much help with generating heart-centered, creative writing. (Trust me; I tried, and the results were disastrous.)
So I’m going to switch up the cadence and content of newsletter posts. Starting today, I’ll be offering My Sweet Dumb Brain readers the following:
1-2 essays a month, available to all subscribers
1 thread a month (here’s an example), for paid subscribers
1 collection of thoughts (a favorite among readers; here’s an example), for paid subscribers
Essays like this one will also include extras for paid subscribers at the bottom of each post, something I’ve offered since the earliest days of this newsletter: related links, recommendations, personal updates, and an exercise for the week ahead.
Adopting a slower pace should give me more time to generate thoughtful essay ideas and make sure that the writing I’m delivering is well-researched and thought-out. It will allow me the space to better connect to each topic, so I can confidently share authentic, insightful thoughts.
Maybe best of all, making these changes will allow me to be kind to myself: a core tenet of this newsletter. I’ll get paid for the work I’m doing (work that AI can’t replicate, at least not yet!) and give myself more opportunities to recharge. I’ll be able to show up here with joy, not apprehension.
I’m thrilled to be able to continue to offer essays to all readers. For as long as I write My Sweet Dumb Brain, that won’t change.
What does it say about me if I trust a robot’s advice more than I trust my own instincts?
This was perhaps the biggest takeaway on my long walks. I’ve been feeling a bit lost lately — with the newsletter, yes, but in many other areas of my life, too — because I’ve forgotten what it means to trust my own intuition. Somewhere along the way, I stopped believing in myself, my ideas, and my capabilities. I doubted why I had this newsletter and a handful of other jobs in the first place.
In the midst of all that doubt, I looked to others for guidance. But no one — especially not an insentient chatbot — knows me as well as I do. If I look back at the prompts I was giving ChatGPT (“Can you give me a different version of that schedule, one where I have time for a walk outside?”) it’s clear I already had the answers I was looking for.
The world is full of people who doubt themselves. It’s also riddled with people who don’t. And while I’ve always appreciated the value of humility, I’ve also seen the many ways that my own lack of self-confidence has held me back.
During my break, I could have chosen to step away from writing this newsletter, just like I’ve stepped away from other creative opportunities. I could have told myself that I’m not offering anything of value to this world; that artificial intelligence is headed for my livelihood anyway.
But I didn’t. Instead, I decided to do what I haven’t done in a long time: to believe in myself and my abilities. To say yes to something that matters. To figure out how to keep creating, in a way that serves me and this amazing community of readers. To know that to be human is to be imperfect, and that imperfection is beautiful.
I’m excited about what’s ahead. I hope you are too.
xoxo KHG
this has given me less shame about asking chatgpt for workouts at the gym and being impressed with how helpful its been lol
I moved to bi-weekly newsletters two years ago (sometimes I even send them once every three weeks!) and no one seems to notice. As a reader, I also appreciate the slower pace of some newsletters - it can be hard to keep up with all my favorites, like yours! All of which is to say, I support and approve the move to fewer, more thoughtful newsletters.