My co-founder and I are building a platform that affects people's economic opportunities, and we keep debating what 'fair' means in our algorithm design. She thinks we should optimize for merit-based outcomes, but I worry that just reinforces existing inequalities, and we can't agree on what we'd choose if we didn't know our own positions in society.
More people experience this than they realize.
Designing systems that affect others' opportunities reveals fundamental disagreements about justice and merit.
“Where Are You with Rawlsian justice — the thought experiment?”
If this sounds familiar, the Library can help you find the bigger picture.