Recent studies explored user interactions with chatbots, particularly focusing on emotional conversations. Participants were divided, with some engaging in open discussions and others following specific guidelines. Despite the differing structures, results indicated no significant variation in user experiences between text-based and voice-based chatbots. A separate analysis of over 3 million conversations with ChatGPT highlighted a surprising trend: emotional discussions are infrequent among users, suggesting limitations in how people choose to engage with AI on a personal level. These findings have yet to be peer-reviewed, leaving their insights preliminary at this stage.
In a comparison between open conversations and those guided by specific instructions, the researchers found little variance in user engagement regardless of the chat format employed.
An analysis of 3 million conversations revealed that, contrary to expectations, emotional exchanges with ChatGPT are quite rare, highlighting potential limitations in its usage.
Collection
[
|
...
]