Some of you might vaguely remember a study I did a year and change ago, measuring an allegedly ChatGPT-generated wordplay against some human-generated alternatives. I’m hoping enough time has passed that such memories are vague rather than specific.
It’s time to gather the concluding data! If you’d like to help do so, then go to this very simple poll and decide which of these is your favorite. Do not look up which of these entries was ChatGPT (as mentioned in The Journal of Wordplay #1-2)—the idea is for this to be a blind taste test.
This test is meant to gather 500 responses, which should be enough to get reasonably accurate results for a single question—and factor out the few respondents who might remember the earlier studies too well. Depending on how quickly the responses come in, they’ll most likely be reported in issue #6 or #7. ’Til then!