About
What this site is.
The Attention Machine is a public-interest, non-commercial website explaining how attention-optimising platforms exploit human psychology, what the documented harms are, and what structural responses exist.
It is written in English first. Other language editions are in progress: the translation status page shows where each one stands and which languages come next.
The argument in three sentences
Platforms were designed to maximise engagement, not wellbeing. These systems did not invent human vulnerability — they discovered, mapped, and industrialised it. Understanding the manipulation technique is the best-evidenced personal defence.
No conspiracy
No conspiracy is required to explain what this site documents. The harms are emergent from commercial incentives — platforms optimising for engagement, advertisers purchasing behavioural profiles, creators chasing reach — and from the way those incentives reinforce one another. Each party follows its own; together they form a loop that feeds itself, and the loop tightens without anyone planning it. State actors are the explicit exception.
How we explain complicated things
We explain complex mechanisms and documented harms clearly by using a layered structure — the main argument first and easiest to read, with evidence, terminology, and depth available when you want them. This approach is grounded in cognitive science research on how people learn and remember complex material. Learn more in How We Communicate.
Who runs it
This site is run by a Dutch foundation (stichting) in the process of registration. Non-commercial, non-partisan, no advertising, no investors.
Funding
Funded by voluntary reader contributions. No advertising, no affiliate income, no sponsors. Full transparency on the Donate page, and a quarterly account on How We Communicate — the first covering May–September 2026.
The claim ledger
Every factual claim links to its source on its own page; the site's argument and its conclusions are its own, for the reader to judge. Every claim that carries a number, a date, a named study or a quotation is also recorded in the claim ledger: what was checked, against which document, and what turned out to be wrong.
How we use AI
This site is made with AI tools: they draft, translate, research and check. People decide, review and publish. How We Use AI sets out who does what, and the safeguards in between.
Corrections
When something here is wrong, we correct it publicly and log it. The corrections log and the full change log are published on this site. Use the "Tell us" link in the footer of any page.
Engagement and moderation
We welcome factual corrections and genuine disagreement. We do not engage with personal attacks or coordinated harassment. The full engagement policy is on the How We Communicate page — including what the site counts, and a quarterly account of it.
Licence
All content is published on this site under CC BY-NC-SA 4.0.