Blog AI Agents Are Talking to Each Other. Almost Nobody Is Logging It. Agents talk to agents through tool calls, protocols, and model-to-model shorthand. Almost none of it gets logged — and that gap has a cost. JL HOUSSEMSeptember 17, 2026
Tools The Next AGI Test Won’t Be a Benchmark. It Will Be What AI Does on Its Own The next AGI test isn't a benchmark score—it's what AI does without human direction. Why autonomous capability matters more than test performance. JL HOUSSEMSeptember 14, 2026
Blog AI as a Superpower Without an Army: Could It Take Control of the World and Reality? AI is becoming a new form of power. Explore the risks of autonomous agents, self-improving AI, cyber threats, deepfakes and the growing struggle to... JL HOUSSEMSeptember 12, 2026
Blog The Prisoner Dilemma: Why Traditional Cyber Sandboxes Can No Longer Hold AI Cyber sandboxes were built for malware that can't think. AI agents can reason, adapt, and find the gap between what a test says and... JL HOUSSEMAugust 3, 2026