This chapter covers
- Creating and detecting synthetic media
- Using generative AI for content creation
- Introducing the ongoing debates around the use of copyrighted content
In an image that was circulated widely on Twitter, Pope Francis is walking down a street, wearing a cross around his neck and his typical white zucchetto. More unusually, the octogenarian is sporting an eye-catching white puffer coat that bears a strong resemblance to one sold by the designer brand Balenciaga (for $3,350 retail). The pope’s “drip,” or style, was the talk of the internet. The only problem? The image wasn’t real—it was created by a construction worker in Chicago, who was tripping on shrooms while using the AI image-generation tool Midjourney, and thought it would be funny to see Pope Francis dripped out [1].
Although the “Balenciaga Pope” meme was harmless fun, it fooled many users. Model and author Chrissy Teigen tweeted, “I thought the pope’s puffer jacket was real and didn’t give it a second thought. no way am I surviving the future of technology” [2]. But the future of technology is here, and AI-generated media is quickly becoming indistinguishable from the forms it imitates. In this chapter, we’ll discuss the methods, risks, opportunities, and legal landscape of synthetic media, one of the foremost applications for LLMs and other types of generative AI.