First, can you tell us a little bit more about the Fandom platform?
“We go across a wide variety of different types of media, our main of course being Fandom,” says Wartel. “A lot of people have been on it even if they’re not necessarily aware. If you want to look up someone in the entertainment space like, say, Mark Hamill, you go to Wikipedia. But if you’d like to look up someone like Luke Skywalker or Leah or R2D2 and see that sort of biography and richness of information, then you come to Fandom and you can see that sort of content right on our website, created by the users that are most involved in those fictional universes. The vast majority of our content is created by our users. And we aim to provide the platform, the tools that allow them to create that content.”
Wartel held a session at TMRE on “Understanding and Managing Perceptions of AI and ChatGPT in a User-Generated Content Community.” It explored perceptions of Generative AI tools through findings collected across a series of studies conducted within the Fandom community. This includes community perceptions, media influences’ shaping of users’ understanding and expectations, and the impact of naming and messaging strategies on users’ acceptance and engagement with tools and content that use AI.
“When it comes to that level of textual information, helping to improve it, or even helping to create it that much quicker, through the use of tools like ChatGPT or other AI is incredibly exciting to us in this space,” observes Wartel. “That is not always a perception that is shared by our users. There’s a bit of concern just like many industries including our own that it may not be a tool so much as potentially a replacement over time. That’s something that we’ve had to contend with as we roll these things out. We have followed up with our users in a variety of different ways. Starting initially with a survey of our core user groups, what we call our editors, our admins. We asked them some questions about their opinions including on AI and their tool set. We didn’t find universal acceptance.”
It feels like you’re close with this group and they’re close with you and you understand each other, but we’ve got this issue here around AI. So how are you solving it?
Wartel continues, “We brought some of those administrators and editors to an event that we call community connect, which this last year was in Atlanta. We ran a series of focus groups. As we talked with them at that event, we did pre and post testing on AI. And one of the things we found in addition to all of the wonderful findings from the focus group itself was that communication really made an impact. We saw the neutral group remain about the same. But those negative reduced by almost half. And we saw increases of up to 10% just from that weekend in terms of those that were now for AI amongst the exact same group of people. How we talk about it, how we inform them can have a real impact. There’s an uphill battle when it comes to that umbrella concept of AI, but not necessarily the pragmatic things you can do with it.”
“We’re at a point where we need to decide not just as a company and probably as an industry, how to discuss this in a way that consumers in different spaces will embrace it,” he adds. “Our work is probably limited to the UGC space, or at the very least the consumer space. Speaking to investors or some other group like that, they get excited about the buzzword that is AI, but it’s not having that impact with our consumers in quite the same way, at least in our space. So we set about doing a name test.”
See the full video interview from TMRE, as Adler and Wartel discuss the name test and attribution testing of the Fandom research study on AI and more.