Why AI-generated slopaganda works (www.machinesociety.ai)

Recently, President Trump took to Truth Social to share a series of AI-generated images, dubbed "slopaganda," which portray him in absurd scenarios, such as attacking the Canadian Prime Minister or wearing military uniforms despite his known draft avoidance. While many may see these images as mere trolling, research indicates that AI-generated content can significantly influence public belief even when labeled as artificial. This effect occurs at a subconscious level, where false images can firmly plant incorrect beliefs that persist despite later corrections. The implications for the AI and machine learning community are profound, as studies demonstrate AI-generated imagery, including deepfakes, can be surprisingly persuasive, often rated as more trustworthy than human-generated faces. As users increasingly engage with AI, their ability to distinguish fake news diminishes, even as their confidence in spotting it grows. Moreover, AI chatbots prove adept at swaying political opinions and potentially addressing conspiracy theories through rapid, rational arguments. However, a challenge remains: harnessing these persuasive capabilities for positive outcomes lacks a clear business model, raising ethical concerns about the misuse of such technologies in misinformation campaigns.