Hacker News Reader

There's value in the converse of that data point, though, right? If someone made rent on April 1st (and maybe even more importantly May 1st), that might be an even stronger signal that they'll continue to do so without interruption.

View on HN

The big differentiator is going to white collar jobs. There will be others, but it doesnt take an AI to figure out who is at risk.

The UK government announced and started support for pay (income tax) as you earn employees a while before support for the self employed/contractors, so maybe the AI prefers standard salaried employment above all else.

> maybe the AI prefers standard salaried employment above all else.

Is it wrong for AI to have the same biases as the general population?

Depends on how common each is. It's a signal that puts people in the top or bottom X% of renters. If knowing someone is in the bottom 90% of renters isn't that useful, then knowing someone else is in the top 10% is, yes. If the threshold is closer to the median, then both sides are closer in usefulness to each other.

I'd be pretty surprised if making rent in April was really at the 90th percentile, though.

Reply on news.ycombinator.com