I'll highlight unusual AI insights observed, discuss insights from a former Google employee on the prevalence of these features, and emphasize a well-known anecdote illustrating the importance of considering human susceptibility in large-scale public experiments.