HumanBehaviorByBjork [any, undecided]

  • 11 Posts
  • 2.25K Comments
Joined 5 years ago
cake
Cake day: July 26th, 2020

help-circle




  • that is not true in the slightest. that is a made up number. i’m not merely reacting with disbelief, i am disbelieving, and i think such claims don’t stand up to even a bit of critical investigation. Let’s do that:

    “have interacted with ai” could mean almost anything. what it doesn’t mean is “interact [present tense] romantically with ai.” the link posted is a pop science site, not a study, and the link that that statistic links to is an outlet for press releases literally called PR Newswire that has a release from an institute at Brigham Young University, the Mormon university that in a civilized nation would not be an accredited institution. Searching based on the title turns up the report here: Counterfeit Connections: The Rise of Romantic AI Connections and AI Sexualized Media Among the Rising Generation. Reading the study shows the method was online survey. I don’t think it’s too much of an assumption to make that this would bias the sample population towards people who have “interacted” with AI chatbots. The specific question that number comes from is "Have you ever chatted online with an AI system or downloaded an AI chat app that was meant to simulate a romantic partner?” so already the question is phrased in such a way that it could easily be misunderstood (have you interacted with an online chatbot [of any kind]), and covers any interaction whatsoever.

    Given these facts, I don’t think this headline statistic is worth crediting.