Insights
Written to be read.
Long-form analysis on artificial intelligence, security and strategic competition from the Camulos team.
Featured
BLUEDOT Technical AI Safety Puzzle #1
Watermarking a model's internal activations
Read the essay 01 Sovereign Chinese AI is a strategic reality Strategy 02 ICLR An Overview of Subliminal Learning ↗ AI Security 03 The Model Provenance Problem AI Security 04 Agents Everywhere AI Security 05 Patterns, Not People: Personality Structures in LLM-powered Persona Agents ↗ AI Security 06 LLM Killswitch AI Security 07 The Père David's Owl AI Security 08 Hybrid Autonomy Sabotage: Innovation Anchored in Tradition ↗ Strategy 09 MoD Data Breach: What Next? ↗ Cyber 10 Voice Cloning @ Scale AI Security 11 Dynamic Serverless Honeynets on AWS Cyber 12 Serverless T-Pot Honeynets on AWS Cyber