Plain-English books on AI, risk, and governance.

A durable way to think about any AI system, across data types and tasks, so a new model or paper stops being a mystery. Plain English, no hype.

One durable way to think about any AI system: three questions - what data, what task, which method - and the math under every model, from a line to an agent. Plain English, no hype, no code.

A curated, no-hype path to actually learning AI - the courses, books, and explainers that teach how it works, from before the hype.

Three moves to collapse twenty AI governance frameworks into one backbone you can actually scale.

Governance that scales is not a longer checklist. Three moves: govern to a common spine, govern with the unit that scales, govern as a system.

A short, board-facing guide to AI risk: five questions to take into the next board meeting, through the financial sector's risk lens.

The guide for the second line: how to run the AI risk-management system day to day, built around six questions to put to any first line, where risk hides, and proportionality as the one tool to cover a portfolio you cannot validate in full.

You have been handed AI risk, and the checklists map to everything and manage nothing. The answer in six questions, read from the second line's seat and drawn from the AIRG.

How to judge a firm's AI risk management from the other side of the table, as a regulator and supervisor.

A few simple questions that tell a supervisor whether a firm's AI risk management is real.

AI risk management as 3 questions, not an ethics poster: what do you have, how are you managing it, who's accountable. From ML to agents, in plain English.

You cannot review every path an AI agent might take. Govern it like a bonsai: cut the branches that must never happen, wire the rest, check with your own hands.

A curated, no-noise starting path for AI risk and governance: six short reading lists by area, each reading with a note on why it's worth reading.

Four agent patterns, with runnable code, to build AI investing tools grounded in real data and auditable calculations - not hallucinations.

Why an LLM on its own makes up your portfolio numbers, and the architecture that fixes it: buns not meat, the complexity ladder, four agent patterns. No code.

Why an LLM cannot forecast on its own, what the real forecasting tools are, and why the rational way to use AI here is an agent that drives those tools. No code.

How generative AI and agents actually work, where they help in finance, and the checks before you let them act.

How generative AI and agents work, where they help in finance, and the checks that follow, in one short read.