#131: How AI might let us be more human.
A mother-daughter story of Pepto Bismol, ChatGPT, and what really matters.
The Luminist is a reader-supported publication that illuminates the pain, the pleasure, and the paradox on the path to technicolor living. Subscribe below to receive posts about how loss teaches us to get the most out of life (along with silly gifs) in your inbox every Saturday.
"I think you can eat regular food tomorrow… if ChatGPT agrees."
I was texting Kendall, curled up in her dorm room four states away, just after surviving her first-ever bout of food poisoning — that special hell that reduces grown adults to whimpering puddles wondering why existence must be so cruel.
During the worst of it, we had texted back and forth — me offering practical advice (suck on ice chips to stay hydrated), battle-tested wisdom (bring a ratty blanket into the bathroom so the floor isn't so miserable) and moral support (this is terrible and will be terrible for a bit. Hang in there, I'm here!).
Then during a lull in the action, she texted: "ChatGPT says if I want to try Pepto Bismol, to start with half a dose.”
ChatGPT?
She sent a screenshot of her conversation with the AI, and for a second, I felt a twinge of... Irrelevance? Jealousy? The digital usurpation of my maternal throne?
Isn't your mom supposed to be the font of all wisdom, the remedy dispenser, the advice giver here? Am I being replaced by an AI?
Then I caught myself. What do I actually know about food poisoning? I've survived my own stomach tsunamis a few times, but were my saltines-and-ginger-ale remedies actually medically sound? Or just what my mother told me when I was throwing up in 1986?
I smoothed my ruffled feathers and texted back: "Well let's go with ChatGPT here. There's no way I know as much about this as an AI. We'll rely on GPT for the medical advice and I'll do what I do best: the 'oh my poor baby' part of being your mom."
This whole episode got me thinking about a book I read recently: Being Wrong by Kathryn Schulz.
Turns out, we humans have an "unstoppable drive to tell stories that make sense of our world." But here's the kicker — we often don't have nearly enough information to build those stories accurately.
Does that stop us?
Not even close. Our egos barrel straight ahead like a monster truck over a speedbump, becoming what scientists call "confabulators": we make stuff up, then convince ourselves it's true.
Back in our cave-dwelling days, this quick theorizing was essential:
That rustling bush = deadly predator or prey for our breakfast?
That colorful berry = delicious snack or slow, terrible death?
That approaching stranger = friend or foe?
This ability to make a snap judgement without nearly enough information apparently kept our ancestors alive well enough that it was passed down to us. The problem is, in our modern lives it causes us to spin elaborate theories based on tiny scraps of information, then treat them as gospel.
Don’t worry, it gets worse. Schulz explains that we do indeed have an inner fact checker who verifies the fantastical tales of our own inner storyteller against our senses and memories. But that fact checker is disproportionately influenced by any initial evidence we encounter. In other words, the first perspective we receive about a situation often becomes our baked-in belief on “how it really is”. Then we become unbudgeable, convinced we're right even when evidence mounts to the contrary.
That’s how I became an instant medical expert the moment my kid got sick. Or how we all became virologists during the pandemic. Or economists during the recession. Or constitutional scholars during election season.
So what can we do about it? Recognize that we will be wrong, regularly. It’s impossible not to be.
Why is it so hard to say "I don't know"?
Why was my knee-jerk reaction to fight a robot that can read the entire internet, rather than graciously cede to its obvious expertise?
Here’s what I’m starting to see: in our western society, we equate knowledge with value, expertise with worth. And so admitting uncertainty feels like admitting defeat. As if not knowing is a personal failing rather than the default human condition.
I see it in myself constantly. When asked for my opinion about a topic I know very little about, I can come up with a point of view in two seconds flat. I think about my kids — though I’ve been through young-adulthood, I have very little understanding of what it’s like to be 20 in 2025, with TikTok and smartphones and photo filters and AI-generated images. But I have plenty of advice to give.
We convince ourselves we know because the opposite, uncertainty, is uncomfortable. Because saying "I don't know" feels like leaving a vacuum others will rush to fill. Because we've built entire identities around being the knowledgeable one, the capable one, the one with answers.
But what if our human superpower isn't knowing everything?
As Kendall lay in bed rehydrating, we FaceTimed. She showed me her AI conversation, including ChatGPT's offer: "Would you like me to help you make a plan for the next 4-6 hours?"
Well, damn. That's actually smart. And far more comprehensive than my scattered advice based on the three times I've been through this myself. (Gatorade! Dry toast! Tylenol for the body aches!)
A few days later, when I mentioned to Kendall that I was feeling bashful letting AI do my mothering for me, she jumped to my defense: "ChatGPT doesn't have the warm and fuzzy," she insisted. "I don't trust it. YOU got the credibility. You've been through it."
And there it is — the real human superpower: connecting through our shared uncertainty, and caring for each other like only living, breathing beings can.
I'm no longer threatened by having an AI in the mix.
I'm grateful for the better information. Let ChatGPT handle what it does so well: organize vast amounts of knowledge. Then I'll do what I do best: consume it, then open my mind and decide what to do (or what to advise Kendall to do) next.
There's wisdom in questioning our theories, gathering better data, and revising our position. There's strength in acknowledging our limitations.
Kendall will still call me first when disaster strikes. Not because I have all the answers. But because I bring something no AI can: the messy, imperfect, deeply human ability to sit in uncertainty, to laugh when things are terrible, to support without being able to fix.
Especially now that we have AI at our fingertips, that connection and camaraderie are worth more than 1,000 ‘right’ answers any day of the week.
To not knowing it all,
"Here’s what I’m starting to see: in our western society, we equate knowledge with value, expertise with worth. And so admitting uncertainty feels like admitting defeat. As if not knowing is a personal failing rather than the default human condition."
As someone who has worked in US public education for the last twenty years, I have both seen this action in others more times than I can count and done it myself more than...several times. Especially as I moved into leadership positions. The urge to come up with an answer rather than say, "I don't know, but I'll find out," is a hard habit to break. Thanks for pointing out that it is what makes us human.
Love this piece!! So relatable, and helps explain the wild things I say for the sake of having an answer.