This experiment was not specifically about whether a kid would wait for the second marshmellow or not (which would be delayed by 20+ minutes), nor whether they would play with the roomful of toys, but to see how they grew up. The real test was to catch up with the adults and see how ‘successful’ they’d become. The experimenters found that those children who waited for the second marshmellow achieved higher grades and had more ‘successful’ better-paying careers.
It’s the concept of delayed rewards vs immediate rewards and is prevalent in the world of machine learning.
This experiment was not specifically about whether a kid would wait for the second marshmellow or not (which would be delayed by 20+ minutes), nor whether they would play with the roomful of toys, but to see how they grew up. The real test was to catch up with the adults and see how ‘successful’ they’d become. The experimenters found that those children who waited for the second marshmellow achieved higher grades and had more ‘successful’ better-paying careers.
It’s the concept of delayed rewards vs immediate rewards and is prevalent in the world of machine learning.