Back to Adam Grant #4 (on the secrets of strong connections), with Adam Grant.
Why Don't Ideal Traits Predict Speed Dating Matches?
Paul Eastwick and Eli Finkel have done studies where people fill out a questionnaire before a speed dating event. Machine learning algorithms find zero predictive utility from anything that people say about themselves or what people claim to want in other people. Compatibility develops when people spend time together instead of matching beforehand.
Adam Grant #4 (on the secrets of strong connections) ·
From the transcript
We can't talk about this without talking about dating. Because the data on matchmaking algorithms, we could put an entire industry out of business immediately. And Monica, this may be more relevant to you than to Dax. I love these studies that Paul Eastwick and Eli Finkel and their colleagues have done, where they have you go in before, the best ones are before a speed dating event, and you fill out a questionnaire with everything you could think of about you, everything you could think of that you're looking for in another person.
And then they have a bunch of machine learning algorithms that use AI to try to figure out who's going to match. And There is zero predictive utility from anything that you say about yourself, either being similar to or complementary to anybody else you've met. And also what you claim you want in other people has no bearing on whether you click with