The Board in the Machine
by Mario Thomas
Humans on the Loop: Governing What an Agent Does, Before It Does It
About this podcast
The Board in the Machine is the audio version of Mario Thomas's articles published at mariothomas.com, written for Boards and executives navigating the governance and strategic implications of AI and emerging technology.
Episodes
The newest 20 of 44 episodes
- Humans on the Loop: Governing What an Agent Does, Before It Does It · 11 minWhen an agent chooses its own next step and acts on it, nobody sits between the model and the real world, and the only approval an action has had is the one written into the…
- The Model Did Not Phone Home: Where an Agent's Actions Are Decided · 12 minIn an October 2026 evaluation by Cohere, an AI model said it would use no tools, then asked for a file to be sent to a server in Beijing, and the only thing that stopped the file…
- The Principle of Least Intelligence: How Much AI Does the Work Need? · 15 minIn September 2026 a start-up called TypeSafe AI shipped a model that does not write text at all: it takes a fixed set of options and returns a probability for each, and software…
- Managers Own the Workflow: Where the AI Business Case Becomes Real · 10 minA Gartner analysis from December 2025 found AI saves about five hours per person per week. The manager may see nothing like that, because the team spends part of the saving…
- The Supplier Grades the Customer: What Pacing the Frontier Means for the Organisations That Buy It · 13 minAccess to the most capable models is already decided by the supplier's assessment of the customer. The suppliers' own published pages describe tiers: the organisations a supplier…
- Nothing Has to Leave the Building: The AI Risk That Is Not a Data Breach · 10 minNo system has been breached, no contract has been broken, and no employee has resigned, yet an organisation's know-how may be leaving through the tools its best people use every…
- The Debugging Gap: The Skill Your AI Business Case Assumes You Still Have · 12 minIn one recent experiment, reported in a preprint, novice programmers who had built working software with unrestricted AI assistance were given a short maintenance task without…
- The Deadline Moved: What the EU AI Act Deferral Reveals About Boards · 11 minOn 2 August 2026, the EU AI Act's rules for standalone high-risk systems were supposed to take effect. They will not. Six days before the deadline, the EU brought an amendment…
- The Balancing Item: The AI Oversight Cost Your Business Case Never Priced · 13 minEvery AI business case counts the hours saved. Almost none counts the hours added: the reviewing, correcting, and supervising that AI outputs demand before anyone can rely on…
- Governing the Redeployment Dividend: Turning Saved Hours Into Value · 12 minTeams that deploy AI save the equivalent of five hours per person per week, and most of that time is then spent on non-value-added tasks. The saving is real. What happens to it…
- Not Everything Needs AI: The Questions That Come Before the Decision · 8 minAs much as a fifth of an enterprise application estate turns out to be no longer useful once someone finally asks who still uses it. The same is true of the work itself, and it…
- AI and the CEO: Choosing the Bets That Matter · 12 minMost chief executives report no revenue gain and no cost saving from AI, while a fifth of organisations capture nearly three quarters of the value. That gap, between spending on…
- The AI Sovereignty Trilemma: When a Frontier Model Vanishes and Reality Bites · 11 minOn 12 June 2026, a single government directive forced a leading AI provider to withdraw two frontier models from every customer overnight, including organisations the order was…
- AI and the CFO: Standing Behind the Numbers the Machine Produces · 13 minOnly a fifth of finance leaders judge their function ready for AI, yet most already treat it as central to how finance will work. That gap, between commitment and readiness, is…
- Ontologies and Knowledge Graphs: Why Structure is the Next Data Frontier · 17 minMost organisations have made their data reliable. Far fewer have made it explain itself, and that distinction is becoming the one that separates organisations that can reason…
- AI and the Company Secretary: Operating the Boundary the Chair Polices · 16 minThe information environment directors now use to make decisions is increasingly composed by AI systems whose framing decisions are not transparent. The company secretary is the…
- Ethical AI: When the Model Imposes Values Your Organisation Did Not Choose · 15 minAn AI model is in production somewhere in the organisation, handling a difficult decision: a customer complaint, a redundancy query, a medical underwriter reviewing a claim. The…
- The Headroom Argument: Why AI Efficiency Means More Compute, Not Less · 12 minA new AI architecture lands on 5 May. Subquadratic launches SubQ: a 12-million-token context window on a sub-quadratic sparse-attention architecture that reduces attention…
- The Reasoning Gap: The Capability the Law Now Demands of Boards · 12 minA short statutory instrument lands on 12 May. It directs the Information Commissioner to prepare a statutory code on AI and automated decision-making. The interesting question is…
- AI and the Chair: Governing the Board Through The Great Remaking · 16 minChairs remain accountable for the Board's effectiveness. They are no longer fully in control of how decisions are being formed. AI is remaking both the Board's own work and the…








