My AI principles

I’m using AI more & more…. I’ve come up with 3 personal ‘AI principles’ I’m using to shape my usage with some ethical guardrails.

They are:

1. AI tools never pretend to be me.
2. I never sell the output of AI tools
3. I never use AI to make a decision that effects someone other than myself.

Here’s the basic reasoning for each one.. they build on each other:

Principle 1 comes from maintaining trust.

If I send a message or email, everyone must know it came from me. The human. If any of my AI systems ever interact with other people they will identify themselves as AI systems working on my behalf.

Principle 2 come from maintaining authenticity

I might use AI tools to help me build stuff, achieve a project goal, or find information… but I’ll never sell the direct output from an LLM. If you see a concept I’ve invented, words on a slide or read an article I’ve written… I came up with the concept, i choose the words & wrote the article etc.

Principle 3 comes from maintaining accountability.

I am accountable for any harm caused by my use of AI. Limiting the scope to decisions about myself helps protect the people around me from harm. This principle is intended to be broad and push me to consider knock on effects.

Final thoughts

I’m starting from the best base I can. I’m using open source models runnng locally. No cloud providers like OpenAI or Anthropic.

I’m using solar power from the array on the shed as much as possible. Around 80% at the moment.

I’m sharing my adventures with tools like Hermes, both the pros and the cons.

All in… I want my use of AI to be positive for myself, my commmunity and society.

What do folks think? Do you have personal principles for AI usage? What are gaps in my principles? All feedback is welcome & appreciated. :)