Under research

What comes next.

Topics currently being studied. The order shows editorial continuity, not a committed publication date.

01 AI model evaluation Benchmarks versus real-world performance, LLM-as-judge and red teaming. Research

02 RAG Architecture, retrieval, evaluation and the limits of grounding. Research

03 Hallucination Why it happens, what reduces the problem and what RAG cannot solve. Research

04 Fine-tuning and adaptation When to adapt a model, when to retrieve context and when to do neither. Pending

05 AI and work Evidence on productivity, augmentation, substitution and skilled work. Pending

06 AI geopolitics Chips, manufacturing, energy, export controls and technological sovereignty. Pending

Browse published series →