How we work.
We read primary sources. We also use AI to supplement our research and provide editorial support. That's what it's there for. We try not to use it uncritically, and we will fact-check ourselves regularly.
We are not interested in creating click-bait polemics. Our interviews are deliberately non-adversarial. We are bored by the endless browbeating of thinkers, policy-makers and entrepreneurs by the 'it's not rocket-science is it?' brigade. Yes, it's rocket-science. Sometimes it's a lot harder than rocket science. And like rocket science, if you get it wrong there are consequences. There are tradeoffs. People making decisions make mistakes. Simplicity is only a virtue if it's a guide to action.
We will also make mistakes. Call us on them. We have conflicts of interest. We will call them out when we're aware of them. Call us on that too.
Questions we are asked
How do you decide what to cover?
One test: does it change what the other stories mean? Each weekly edition names the week's single most conversation-changing development and argues for it, then reads the money, the rules and the frontier in that order. Most news moves nothing. A headline is a reading on an instrument, and the first question is always which instrument it is a reading on.
How do you label uncertainty?
Every substantive claim sits in one of three registers, marked in the prose. Established means replicated, documented or directly observable. Probable means the balance of evidence, with the reasoning shown. Speculative is flagged as conjecture. Major analyses carry a section headed what would change our mind, naming the observation that would falsify the reading.
What are the sourcing rules?
We read primary sources. Numbers carry dates and the name of the body that published them. Studies are named; research shows is not a citation. Where a number is an estimate we say whose and how it was derived. Where the evidence is thin we say so before someone else does. The sources a piece rests on are listed at its foot.
Do you use AI to write the publication?
We use AI for research support and editorial work. That is what it is for. We do not use it uncritically and we fact-check ourselves. A person is answerable for every claim, every number and every source in a piece. Nothing publishes as a machine's unchecked output.
How do you handle corrections?
Write to corrections@theaiomicage.com. We publish the correction, date it, and leave it visible. We do not amend silently. If a correction weakens a conclusion we say by how much; if it removes one, we say that too. We will make mistakes, and we would rather hear about them from a reader than from a competitor.
Are the interviews adversarial?
No. We come prepared, we ask short questions, and we let the answers stand. We disagree with arguments and never sneer at people. Guests get accuracy checks, not approval rights: they can correct a misquotation or a fact, and they cannot remove an answer they later regret. Quotations are verbatim, and trims are marked.
How do you handle conflicts of interest?
We have them. Both hosts work in and around the industry we cover. Where a topic touches an interest, one plain sentence at the top of the piece says so. Editorial copy does not name a company either founder owns or advises, and no commercial relationship buys coverage, framing or a guest slot. Tell us when we miss one.
Is any of this advice?
No. The guides map the evidence and the tradeoffs and show how the answer changes with a reader's situation. They inform judgment; they do not replace it. Nothing on this site is individual financial, legal or medical advice, and nothing here is a recommendation to buy, sell or hold anything. The decision stays yours.
Last reviewed