About the guide.

We used AI to organize 414 podcast transcripts, then selected events, recurring topics, arguments, forecasts and terms for this guide.

Summaries were checked against transcript passages. Expand a source to read the excerpt and find the original episode and timestamp. Excerpts may start or end within a speaker’s turn; the timestamp marks that turn’s beginning. Separate excerpts are shown separately, without joining them into a continuous quotation.

The 27 conversation guides draw on a reviewed inventory of 133 transcripts from September 2025 through September 4, 2026. Every episode’s topic inventory was screened for major omissions. Each guide connects ideas across multiple episodes around a bounded question, not an exhaustive record of every related remark. Skilled immigration has its own guide alongside immigration enforcement.

The six featured conversations are editorial starting points, chosen for sustained attention, depth and breadth of subject matter. Their placement is not a measured popularity or airtime ranking. Supporting passages provide context for the arguments and their tensions. The number of sections or excerpts is not a measure of how often a subject appeared.

Sources show the reasoning behind an argument; they do not independently verify every factual claim. Interviews can include company claims, disputed accounts and forecasts. We keep distinctions such as alleged, proposed and predicted where they affect the meaning. Transcript errors can also appear in quoted passages.

This is a selection, not a complete record. AI summaries and speaker labels can be wrong. Charts measure word usage. A forecast is distinct from an observed outcome.

Conversations connect developments and ideas across time. Debates focus on specific contested choices within or across those themes. They compare distinct positions, sometimes made on different dates rather than in a direct exchange. There is no fixed number of positions: several people can share an argument, and one person can support parts of different arguments. The cases need not have equal merit or exhaust the possible answers. Featured links are starting points, not a limit on coverage.

We do not publish comprehensive profiles of personal viewpoints or a prediction scorecard. A selection of attributed arguments is not a complete account of a person’s beliefs. Any future forecast entry must have a verified speaker, the full claim and its conditions, an explicit resolution rule, and evidence appropriate to that rule.

Listening paths use the selected sources behind the current conversation guides, not every mention of a topic. Glossary usage counts measure the listed aliases, not the importance of an idea; missing counts do not mean zero mentions.

Feedback or corrections? DM JM on X ↗