Monthly thinking room
How much responsibility can be handed over to agents before control, knowledge and trust can no longer keep up? This edition shows how companies manage agents, review them and supply them with their knowledge, and where adoption gets stuck at interfaces and fears.
For working with your own AI assistant. How it works
We handpicked 107 podcast episodes released in August 2026 and distilled 321 insights from them. These ten patterns show how companies provide agents with knowledge, organize responsibility and oversight and embed AI in everyday work.
Beam stores its entire company context in GitHub, one repository per customer with readme files and instructions, so that a colleague's coding agent can find the information.
Moody's has one person on the team maintain the agents' context layer each week on a rotating basis; at Statista, whoever creates a data series answers five review questions, which raises answer quality by 15 to 20 percentage points.
Instead of Figma designs, Thermondo's design team delivers finished components and skills that teach agents how to use them; this lets engineering and product management assemble interfaces in the design team's style.
Status: confirmed, 2 new supporting episodes. Zapier is building a marketing brain from which people and agents draw the same current goals and brand guidelines, and Dan Slagen considers at least one full-time position for maintaining it worthwhile; Max Kanat-Alexander of Capital One advises writing the command for running the tests into AGENTS.md if the coding agent cannot otherwise run them itself.
Supported by these conversations
BNY gives each of its more than 150 digital employees its own login, a human manager, onboarding and a performance review, and also measures effectiveness by hallucinations and model drift.
Lexware describes the handover to AI as a mandate made up of five components (goal, scope, tone, escalation, control) and provides more than 40 ready-made mandates on a dedicated website.
In John Healy's client case, every agent has metrics and is kept, improved or shut down; Kate Smaje of McKinsey warns that agents without lifecycle rules keep running as abandonware.
Status: unchanged
Supported by these conversations
Beam achieves around 95 percent accuracy with a low-cost model and a second low-cost model as a checking layer, while the same prompt on the latest model generation lands at about 74 to 75 percent.
An ongoing film production stays with ByteDance's older model Seedance 2.0 because Seedance 2.5 was not better in testing, only longer, and thus produces at 20 percent of the cost.
Haus Cramer Gruppe moves the language model and ticketing system for its agentic expansion stage into its own hosting so that no provider can charge 30 percent more tomorrow; Libra lets law firms choose between models from Google, OpenAI and Anthropic for each task.
Status: unchanged
Supported by these conversations
With a subscription
With a subscription you read all 10 with their sources and download them as a text file.
The complete collection
107 conversations, ordered from newest to oldest.
Mandate instead of individual prompts
The leap lies not in better prompts but in the mandate: goal, a framework with prohibitions, tone, escalation thresholds and depth of oversight turn operating AI into genuinely leading it. This matters for decisions because responsibility and liability thereby remain explicitly with management.
With a subscription
With a subscription you open all 107 conversations of this room with their insights, search the catalogue and download the files.
6 conversations added · none removed · 4 revised
2 confirmed · 7 unchanged · 1 with counter-evidence
With a subscription you see which conversations, findings and recommendations changed.
Compared with V4 · 30 Sept 2026.
