> Third, despite having been provided with a clear definition
of AI companions, some respondents may have conflated
general AI use (such as ChatGPT for homework help) with
AI companion interactions, potentially inflating usage
statistics.
I think the viability of this entire study hinges on how clear they were that they are not talking about ChatGPT, even if they use ChatGPT for things like getting advice.
> Third, despite having been provided with a clear definition of AI companions, some respondents may have conflated general AI use (such as ChatGPT for homework help) with AI companion interactions, potentially inflating usage statistics.
I think the viability of this entire study hinges on how clear they were that they are not talking about ChatGPT, even if they use ChatGPT for things like getting advice.