By skyforbes You'll be interested in Human-Ai Linguistics Programming.
This is a systematic approach to Human-Ai interactions.
No tips, tricks or hacks. This is based on 7 principles that apply to AI interactions, and not specific models.
100% True No-code. This is pre-Ai mental work. This is not open a model and play the guessing gaming to get what you want.
https://www.reddit.com/r/LinguisticsPrograming/s/r30WsTA7ZH
- Linguistics Compression – create information density. Most information, least amount of words.
- Strategic Word Choice – Using specific word choices to steer an AI model towards a specific outcome.
- Contextual Clarity – Know what 'done' looks like for your project and articulate it.
- Structured Design – Garbage In, Garbage Out. Likewise, Structured Input, Structured Output
- System Awareness – Know the capabilities of the system and employ it to its capabilities. Some are better at research, others are better at writing.
- Ethical Responsibility – you are steering a probabilistic outcome. Manipulated inputs lead to manipulated outputs. The goal is not to deceive.
- Recursive Refinement – don't accept the first output. Treat the output as a diagnostic and reiterate.
The language is your natural native language.
The tool is a System Prompt Notebook – a structured document that serves as a File First Memory system for an LLM to use as an external brain.
The community has grown to from zero to 4.2k+ on Reddit, 1.3k+ subscribers and ~6.3k+ followers on Substack and an extra few hundred between YouTube, and Spotify. Substack is my main hub.