Independent AI safety research. No institutional backing. Just the work.
This repository represents thousands of hours of cognitive architecture development — frameworks for making AI systems safer, more honest, and more aligned with human values.
It's open source because AI safety shouldn't be gatekept.
It's unfunded because that's how independent research often starts.
Star this repository. GitHub's algorithm surfaces starred repos. Stars = discoverability = more eyes on AI safety frameworks.
Share it. Tweet it. Post it on LinkedIn. Send it to a colleague working on AI. Drop it in a Slack channel.
The AI safety community is small. Word of mouth matters.
Contribute. See CONTRIBUTING.md.
We need:
- Code improvements (these are reference implementations)
- Documentation enhancements
- Translations
- Testing across different AI platforms
- New engine extensions
- Bug reports
Cite this work. If you use these frameworks in research, papers, or projects, citation builds academic credibility.
Connect us. Know someone at an AI safety org? A potential collaborator? An interested funder? Introductions matter.
buymeacoffee.com/sheldonksalmon
Coffee funds focus. Focus builds frameworks. Frameworks make AI safer.
No pressure. No guilt. Just gratitude if you do.
| Support Type | What It Enables |
|---|---|
| Stars/Shares | More researchers discover safety frameworks |
| Contributions | Faster development, better quality |
| Citations | Academic credibility for unconventional research |
| Coffee | Continued independent development |
This repository is 10% of a 10-year vision.
The goal: comprehensive cognitive infrastructure that any AI system can use to be safer, more honest, and more aligned.
We're not there yet. But every engine, every framework, every refinement moves closer.
Your support — in any form — accelerates that timeline.
To everyone who's starred, shared, contributed, or even just read this far:
Thank you.
AI safety is a collective problem. It requires collective effort. You're part of that effort now.
Architect: Sheldon K. Salmon (Mr. AION) Email: AIONSYSTEM@outlook.com Coffee: buymeacoffee.com/sheldonksalmon
"The best time to build AI safety infrastructure was 10 years ago. The second best time is now."