https://www.nytimes.com/interactive...7366&user_id=c383821527c441214d07ce6e4a6ba12a
In July, professors at Middlebury College released a preprint study in which more than 200 undergraduates were asked to write an essay and take a proctored multiple-choice test.
One group of students had access to A.I. to research a topic and write the essay. The other group didn’t. The researchers found that the A.I.-enabled group wrote longer and slightly higher-quality essays. Their test scores were also higher.
A week later, the participants returned for a similar test, but now both groups worked unassisted. The authors broke down the students’ second-round performance based on how A.I. was used in the first round.
In July, professors at Middlebury College released a preprint study in which more than 200 undergraduates were asked to write an essay and take a proctored multiple-choice test.
One group of students had access to A.I. to research a topic and write the essay. The other group didn’t. The researchers found that the A.I.-enabled group wrote longer and slightly higher-quality essays. Their test scores were also higher.
A week later, the participants returned for a similar test, but now both groups worked unassisted. The authors broke down the students’ second-round performance based on how A.I. was used in the first round.