How to become a 10x ramble-coder

Open original

The article argues that speaking long, unfiltered “rambles” to an LLM captures your full train of thought and design trade-offs in a way short typed prompts can’t. It offers tips like embracing uncertainties, recording group debates, and voicing emotions to give the model enough context to make decisions that match your priorities. Finally, it shows how to set up a dictation app or transcription pipeline to feed those recordings into your coding agent.

  • Rambling long, unfiltered voice memos to Codex (transcribed via Whisper) beats short typed prompts because it captures hidden preferences and context, sharply reducing follow-up questions
  • Recording hours-long debates between cofounders gives the AI "ground truth" on team biases/priorities, effectively making it a third team member
  • Deliberately voicing uncertainties, false starts, and emotional reactions (not just polished conclusions) reveals constraints that let the AI make on-target decisions

Andrej Karpathy’s shift to “ramble sessions” with Codex has turned into a major productivity boost for this writer and his cofounder. Instead of typing concise prompts, they talk through every twist and turn of their thought process—uncertainties, false starts, emotional reactions—into a voice memo. They then transcribe it with Whisper and feed it to Codex. The result: the AI picks up on hidden preferences (like disliking redirects or generic UI) and context that a short written prompt would miss. Follow-up questions drop sharply because the agent already understands the team’s style and constraints. They’ve taken this a step further with multiplayer rambling. During “constructive arguments,” both founders debate technical and strategic choices out loud, encourage disagreement and dig into minor points. Those hours-long conversations become the “ground truth” for Codex, effectively turning it into a third team member that knows their biases and priorities. This method applies to everything from feature design to marketing copy. To ramble effectively, they advise leaning into uncertainties and emotional reactions. Spelling out why one option almost worked reveals constraints. Voicing frustration with wording or broken features gives the agent access to the user’s priorities in real time. This richer input cuts down on back-and-forth, delivers more on-target suggestions, and helps the AI mirror the team’s decision process without constant manual clarification.

Ask about this article