The gap thought leadership can't close

Harvard Business Review ran a piece in March 2026 asking whether AI has ended thought leadership. The argument: AI has made confident, well-written commentary about the future of work nearly free to produce, which means there is now a widening gap between people who talk about how AI is reshaping work and people actually building a response to it.

That gap is real, and it is the reason this page exists in this form rather than as another opinion column. If polished argument is now cheap, the only differentiator left is whether the argument is tied to evidence that can be checked, and whether the person making it is doing anything with it beyond publishing.

Concretely, that means three commitments for everything published under Field Notes:

• Every claim about labour-market impact cites a primary source, not a paraphrase of one.
• When new data changes the picture — for example, an updated Anthropic Economic Index report — the relevant page gets revised, not just added to.
• The commentary stays tied to a working artefact: the Cupel framework itself, which exists to be used, mapped against, and argued with, not just read.

The underlying bet is unfashionable: that in a market flooded with AI-assisted takes, being slow, cited, and revisable is the actual differentiator — not being fast and confident.