Recording a Post While Driving, and an Idea About Voice

TL;DR

• Recorded a post while driving to see a friend I hadn’t seen since 2022.
• The real test was whether transcription holds up with road noise in the background.
• If it does, I can capture thoughts where I actually have them rather than at a desk.
• Idea: voice-based microblogging where comments are voice too.
• With voice-only subscriptions, and recommendations matched on voice attributes and interests.

I recorded this while driving to a friend’s house — someone I haven’t seen since 2022.

The practical test was whether AI transcription copes with a noisy environment. If it does, the whole proposition changes: I can capture things where I actually think of them, in a car, rather than only when sitting at a desk with a good microphone. Authenticity in a voice blog partly means catching the thought when it happens, with whatever nuance and background it comes wrapped in.

The idea

Reflecting on past camping trips, something occurred to me: voice-based microblogging where the comments are voice too.

Text comments on a voice post feel like a category error. If someone’s speaking, replying in their own voice preserves what made the original worth listening to — tone, hesitation, the things text flattens. Navigating with voice commands would keep the whole interaction in one mode.

Two features that follow from that:

• Voice-only subscriptions — follow people whose thinking you want in your ear, not your feed.
• Recommendations matched on voice attributes and interests. How someone speaks carries information about whether you’d want to listen to them at length, separately from what they’re speaking about.

Small steps today, but worth testing whether the recording holds up before building anything on the assumption that it does.