30 minutes with a stranger

One of the problems with their "better / worse" statistics: Bad interactions tend to outweigh good interactions. I think the rule of thumb is that 4:1 good/bad ratio in a relationship is "breakeven" where the relationship will stay neutral; higher than that and things get better, lower than that and things go south.

So if you could talk to a stranger, and there's only a 20% chance you'll feel worse, a lot of people would still not consider it worth the risk.

Makes me think the real challenge isn't just encouraging people to talk to strangers, but designing situations where the expected value of those conversations skews heavily positive