Explaining It to People Who Do Not Get It

I tried, for a while, to explain vibe coding to people at dinner.

It went badly. Not because they were rude — they were kind. It went badly because there is no clean sentence for I type English into a chat window and a working app comes out the other side and I am doing this instead of sleeping.

They'd nod. They'd say "that's cool." They'd change the subject to college football, which, to be fair, I was also happy to change the subject to.

For a while that stung. I wanted them to see it. I wanted them to be excited. I wanted somebody in my actual life to say dude, this is a big deal.

What I learned

Most people don't need to get it. That's fine. That's actually the correct outcome.

• My wife doesn't need to understand the tech to support the person doing it. She just needs to know I'm okay and the trash is out.
• My friends don't need to validate the projects. They need to still want to golf with me.
• My kids don't need to be impressed. They need a dad who's around, not one who's constantly explaining a dashboard.

The audience for the work is not the people at my dinner table. The audience for the love is.

Once I separated those two things, dinner got a lot better. And so did the work.

A small script that helped

When someone asks what I'm doing now, I don't try to explain vibe coding anymore. I just say: I'm building some small internet projects, it's a fun hobby.

That sentence is true. It is also complete. It closes the loop without asking anyone to carry something they didn't sign up for.

Save your explanations for the people who actually want them. There are more of those people than you think — they're just usually online.

The three conversations, and what each one is actually asking

Once I stopped treating every "so what are you working on?" as the same question, this got much easier. There are really only three:

1. The polite question. Someone at a cookout making conversation. They want fifteen seconds and a way back to normal talk. Answer: "Some small internet projects — kind of a fun hobby." Done. Everyone is comfortable.

2. The worried question. Usually family. Underneath "what are you doing?" is "are we okay, and is this costing us money?" Explaining the tech does nothing here. Answer the real question: what it costs, how much time it takes, and what would make you stop.

3. The genuinely curious question. Rare, and you can hear it — they ask a second question. This is the only one that earns a real explanation, and it is worth having a clean one ready.

My clean one, after a lot of failed attempts at dinner tables: "I describe what I want an app to do in plain English, and the AI writes the code. I am basically the person deciding what gets built, not the person typing it."

That sentence works because it does not ask anyone to care about tools. It puts a familiar job — deciding, directing — where the unfamiliar one was.

What I got wrong for a year

I was using explanation as a request. I was not really saying "here is what I do." I was saying "please tell me this is worth doing."

That is a heavy thing to hand someone over a plate of ribs. They cannot give it to you, so they nod and change the subject, and you walk away feeling unseen — when really you just asked the wrong person for the wrong thing.

Validation comes from users and from shipped work. Support comes from the people at the table. Asking either one for the other job is how you end up disappointed in both.

Where the understanding did come from

Online, and later than I would have liked. The AI-builder communities are full of people at exactly this stage, and one comment from a stranger who has hit the same wall does more than an hour of trying to explain yourself to someone who loves you but does not care about databases.

Go find those people. Then let your dinner table off the hook. Both relationships improve immediately.