Wulfy—Speaker to the machines1d ago
Robert Sanscartier

#AI

Realistic AI-generated climate disaster images decrease support for climate action when artificial origin is suspected

Policymakers and environmental advocacy organizations are increasingly using AI-generated images of climate disasters to advocate for climate interventions. Here we show that highly realistic AI-generated climate disaster images do not increase support for climate action

https://www.nature.com/articles/s44458-026-00092-0

#ClimateChange #UpheavalClimate #pollution #ecology #environment #climate

Realistic AI-generated climate disaster images decrease support for climate action when artificial origin is suspected - Communications Sustainability

Support for climate action decreases when highly realistic disaster images are suspected to be generated by artificial intelligence, as shown by three large experiments testing emotional and resistance responses to climate disaster imagery.

Nature
May 20 at 8:26amWeb
Show threadPetra van CronenburgMay 20
@Snoro Good to see a study. It is so easy: if I see an #AISlop image, I don't read the text. I don't believe the message because someone using AISlop produces fakes.
Show thread🔥 Mireille Sillander 🔥May 20
@Snoro I'm glad that people studied it, because it seems obvious to most people even tangentially interested in protecting nature that using AI to promote the cause would be astoundingly hypocritical, but unfortunately, the marketing and managerial levels of any conservation org don't seem to be connected with general audiences
Show threadzombiecide　馬雄May 20
@Mimesatwork Perceived hypocrisy is one aspect, but I think the whole reactance and reduced trust has other aspects. Like, the whole emotional impact itself makes people protect themselves and feel distrustful because they don't want to experience more emotional upheaval brought unto them by somebody trying to make them do things. Most people don't seem to care much about other's hypocrisy if it supports their own stance/comfort zone?
Show threadzombiecide　馬雄May 20
@Mimesatwork And I think one issue with the disconnect is those mnagerial types believing "if people just knew they'd change their ways" when it's systemic issues that make people behave the way that is easiest in their current situation
Show threadPetra van CronenburgMay 20

@zombiecide If you look at the study, it's only about using #genAI pictures for #activism, not other arguments.

I'm not a scientist and don't know activists from such organisations using slop. But what I experience with activists is:
1. Most people don't think that these images are hyperrealistic, they feel something "weird".
2. You only protect what you like or love. For that purpose, you have to know it. Therefore, good nature photography works much better.
3. People get active

@Mimesatwork

Show threadPetra van CronenburgMay 20

@zombiecide if they feel agency/hope. You don't get them active by threatening them! Yes, give them emotions: 'you can do this or that!" Let them feel agency. You can't provoke such emotions with #AISlop nor with doom & gloom!

@Mimesatwork