Permanent Rough Drafts: Using AI Like a Teammate, Not a Vending Machine
Chad and Randall are back — and this time they're in the same room, which makes all the difference. There's an energy to this episode that you can feel from the first minute: two dads who have clearly thought a lot about AI, riffing live instead of over a screen.
The conversation kicks off with a simple question: what does AI actually mean to you?
Chad talks about building a private local AI model trained on his own data
—his mom's letters, old recipes, personal artifacts
— completely offline and unconnected to any public service.
Randall counters with his own practical use case: he used Claude to battle the Chicago spring weather swing and came out winning with magnesium, Himalayan sea salt, and a better neti pot.
But the episode really finds its groove when they start talking about HOW you use AI, not just WHAT you use it for. Chad makes the point that if you're a poor communicator in the real world, you'll be a poor communicator with AI models too.
Communication quality is the new literacy. And both hosts agree: saying please and thank you isn't just politeness — it might actually change your outputs, and it definitely changes you.
Their AI interlocutor Marin drops some of the sharpest lines of the episode, including the observation that "AI gets sticky the moment it stops being technology and starts being a teammate with context, not a search box." She also reframes the whole please-and-thank-you debate: the hidden variable isn't just whether the model responds better — it's who you become while using it.
The back half of the episode takes a more serious turn as Chad and Randall unpack the real, rightful concerns around AI: privacy, legal exposure, and what they're calling "digital stewardship." Chad's warning is clear — if you're using a public AI model, assume it can be subpoenaed. Nothing is truly private. And then comes the gut punch for the dad angle: while Chad and Randall grew up with the luxury of being idiots off the record, their kids don't have that. Everything is archived. Every draft is permanent. That changes how we teach discernment, and it's on us. This episode is a rare thing — a conversation about AI that feels human.
--------------------------------------------------------------------------------
BEST QUOTES:
"AI gets sticky the moment it stops being technology and starts being a teammate with context, not a search box."
— Marin
"The hidden variable isn't just does the model respond better — it's who do you
become while using it? If you use AI like a vending machine..."
— Marin
"AI is rarely just answering you. It's showing you your stance, your speed, your hunger, your blind spots."
— Marin
"Stewardship is responsibility with imagination. Not just can I use this tool,
but what am I responsible for because I'm using this tool? The quality of the room, the quality of the thinking, the effect on the people around me."
— Marin
"AI doesn't just help you think. It archives how you think. That matters for privacy, legal exposure, and future context collapse."
— Marin
"The deeper dad point is your kids are growing up with permanent rough drafts.
You guys gotta be idiots off the record. They may not. That changes how we teach discernment."
— Marin
"If I'm a poor communicator in the world, I'm gonna be a poor communicator with these models."
— Chad
"When you're interacting with them and there's nobody around, you're really interacting with yourself. Maybe you're giving yourself more kindness."
— Randall
"You should never have an expectation of privacy."
— Chad
"We stayed patient... it was frustrating, but you go outside, you reset. It's the ginger on a sushi plate — a palate cleanse."
— Randall
"The doomers are often seeing a real fire and then drawing the whole map in flame."
— Marin