An independent research review institute
AI answers.We ask why.
We read the science of human judgment and AI — neuroscience, psychology, machine learning and economics — and explore what it might mean for work and the people who do it.
Four lenses, one question
What stays human when AI does the work?
Every article starts from the research, and ends with our reading of how much weight it can carry — for now.
Mind
What might the brain tell us about judgment?
Often quoted, rarely read closely. We look at the studies behind the headlines and consider what they suggest — and where they may be stretched.
All Mind articles →Behaviour
How do people judge, decide and trust — alone, in groups, and beside a machine?
Psychology offers some of the most closely measured evidence on how people make decisions. Many well-known findings have also struggled to replicate, so the articles here try to say, in plain words, which ones seem to have held up.
AnalysisThink alone, then talk: what nine doctors who never met can teach every teamAll Behaviour articles →Machine
What might AI do well — and where might people matter most?
How models and decision systems seem to work in practice, as far as the papers and case files show.
All Machine articles →Market
Who gains, who pays — and where does the work go?
Workload, caseloads and labour data: how automation may be changing work that depends on people, and the people who do it.
All Market articles →How we read research
We don't run the studies. We read them closely.
AWHY sits between the lab and everyone else: the public, business and policy. We try to explain what the research suggests, how much it can support, and what it could mean for work.
Primary sources first
Peer-reviewed papers, preprints with methods, official data. The press only for the facts of a case.
Our reading, in plain words
Every article says how much weight we think a finding can carry, and why — in sentences, not scores.
The other side
We make the best case against our own reading, and note what evidence could change it.
Caution where it’s due
We point out when a claim leans on brain scans, small samples, findings that haven’t replicated, lab-only results or vendor-funded studies.
Read it here. Argue about it there.
New articles and the conversation around them, in our LinkedIn newsletter.