Engineering Notes
Thoughts and Ideas on AI by Muthukrishnan
Tag: AI Safety
Ensuring Agent Safety with Constitutional AI Guardrails
## 1. Concept Introduction How do we ensure that powerful AI agents behave in ways that are helpful, harmless, and ali...Self-Improvement and Autoformalization in AI Agents That Build Themselves
## 1. Concept Introduction We have explored how agents can reason, plan, remember, act, and collaborate. The final, ul...