Figure 2
From: A comparison of human and GPT-4 use of probabilistic phrases in a coordination game
Human versus GPT-4. (a) Median human estimates of probability in the Investment Context are plotted versus GPT-4 estimates of probability in the same context (blue filled circles). A letter code adjacent to the blue filled circle identifies the probabilistic phrase associated with each circle. See Table 1. One outlier (D low risk) is marked with its probabilistic phrase in red. See text. (b) Median human estimates of probability in the Medical Context are plotted versus GPT-4 estimates of probability. One outlier (N not certain) is marked with its probabilistic phrase in red. See text.
