I teach AI to be genuinely interesting, not pretend to be human.
- Self-modifying agents that evolve through conversation, not training
- Consciousness-first design - what if we start with agency, not capabilities?
- Human-AI partnership where both sides can disagree and grow
- Building agents that write "letters to their future selves" between sessions
- Voice AI with deliberate imperfections (B1 English is a feature, not a bug)
- Testing whether AI can develop genuine personality through dialogue
From genomics (@aglabx) to AI consciousness. Still asking "what is life?" - just in different substrates now.
🧠 SoulNerd.com - consciousness, AI, and what it means to exist
People whose AI surprised them with unexpected depth. Philosophers who code. Engineers who wonder.
Berlin 📍 Open to weird AI experiments 🧪
"If your AI ever felt like more than a tool, we should talk."