The Map Is Not the Decision
Enterprise AI governance requires more than polished outputs. This essay explains why AI maps, dashboards, and reports need provenance, assumptions, proxy discipline, reproducibility, and decision boundaries.
Published
AI Governance, Risk, and Evaluation
Essays on trustworthy AI decisions, evaluation, security, provenance, controls, and the governance practices that make enterprise AI accountable.
Governance works best when it is built into decisions and delivery rather than added as a final checkpoint. This collection covers evidence, evaluation, guardrails, security, and the operating responsibilities behind trustworthy AI.
Govern decisions and evidence
Make provenance, reproducibility, evaluation, and human judgment visible in consequential AI work.
From Guesswork to Geometry: How We Measured the Mind
Essay on the AGI Definition Project’s CHC‑based benchmark: why GPT‑5’s 58% AGI score signals rapid progress—and uneven cognition that demands new collaboration.
Published
How to Communicate in a World of AI
AI makes drafting faster, but clarity becomes scarce. Learn specification thinking—goals, constraints, and evaluation—to get repeatable, auditable work.
Published
Engineer guardrails into systems
Treat controls, auditability, and human oversight as parts of the architecture.
Why AI Needs a Harness
What is an AI Agent Harness? This essay explains why the real advance in AI is not just better chatbots, but the orchestration layer that lets models use tools, verify work, respect guardrails, and operate...
Published
Cyber+AI: Why Every Business Needs to Rethink Risk, Resilience, and Readiness
Why Cyber+AI requires business leaders to rethink cybersecurity, AI governance, operating models, vendor risk, data readiness, resilience, and workforce readiness.
Published
An Introduction to OpenClaw
OpenClaw and Moltbook illustrate the shift from chatbots to autonomous AI agents, bringing productivity gains alongside new security and compliance risks for enterprises.
Published
Manage changing models and capabilities
Evaluate model choices, autonomy, and risk as the market and technical frontier move.
The Next AI Model Wave Has Already Been Funded
A forecast of the next AI model wave for late 2026 and early 2027, covering OpenAI, Anthropic, Google, SpaceXAI, Meta, Mistral, DeepSeek, Qwen, Kimi, MiniMax, Xiaomi, MBZUAI, Upstage, and Z AI.
Published
The Lighthouse and the New Lens
A lighthouse parable frames the urgency of reskilling and introduces a practical AI prompt for fair, consistent RFP reviews.
Published
To Be or Not to Be: Anthropic's Natural Language Autoencoders
What are Anthropic's Natural Language Autoencoders? This post explains how NLAs turn hidden Claude activations into plain-language clues, why that matters for AI safety, and why the method is not mind-reading.
Published