On September 10, 2019 Gary Marcus and Ernest Davis published a piece at Wired titled "If Computers Are So Smart, How Come They Can’t Read?". They attempted to test GPT-2 using the story by Laura Ingalls Wilder concerning the lost pocketbook. They generated a continuation of the story, but it was not completely coherent. Overall, Marcus and Davis stated that "robust machine reading is still such a distant prospect".

Now, ChatGPT has succeeded in providing good answers to their story questions. However, they might claim that the training set was contaminated, i.e., information about the story, the questions, and the corresponding answers may have appeared in the training set. Thus, they should construct a new set of queries for different children's stories. Also, perhaps they should graciously acknowledge the progress.