Should Pixel 8 AI-edited photos be labeled as synthetic content on YouTube? YouTube is still figuring it out.

Google is diving headlong into the future of AI, which means different parts of the company are bound to collide with each other. Search now generates AI-powered summary results, but flooding the web with AI-generated text will make search faster. much more difficult, for example. This is going to be complicated, complicated and fascinating all at the same time. In general, the parts of Google responsible for moderation platforms (like YouTube and Search) will eventually collide with the parts of Google that help people create content (like Gmail and Bard.) You can simply group them into two lists, draw a line from one to the other, and predict a long list of interesting problems that will inevitably arise.

This week’s Google AI Collision(TM) takes place between the Pixel 8, whose Best Take feature uses AI to let you swap facial expressions in group photos to get “perfect” photosand YouTube, which has just been launched the first version of an AI-generated content policy which requires “creators to disclose when they have created edited or synthetic content that is realistic.” The blog post specifically cites content that “realistically depicts an event that never happened” as requiring labeling, and notes that penalties for consistently failing to label realistic synthetic content can include removal and demonetization of content.

Well! THE Pixel 8 with Best Take and Magic Editor in Google Photos can very clearly generate realistic photos of events that never happened. Specifically, Best Take lets you create moments between people that never happened before by choosing between multiple facial expressions, and Magic Editor can simply remove elements from the background. It’s trivial to make it seem like people are looking at the same thing, or different things, or reacting to each other in ways that never happened – a Pixel 8 photo of Taylor Swift and Travis Kelce n It’s potentially just a few clicks away from telling a very different story from reality. The same goes for a Best Take photo of Donald Trump in the upcoming election, for that matter. Google’s Pixel 8 ads are a great example of how simple it is:

SO! Should Pixel 8 photos edited with Best Take or photos edited with Magic Editor in Google Photos be tagged on YouTube? We asked, and YouTube spokesperson Jack Malon said context would be important and offered some preliminary advice.

“The goal of this update is to help viewers be better informed about realistic but edited content, without punishing creators who use AI,” he says. “If you use Best Shot to choose an image where everyone has their eyes open, that’s something you probably could have done yourself by taking more photos (and probably not something you need to disclose). But if you use technology to make it appear as if you were in a place you weren’t actually there, or if you edit a photo of a historical event, you will likely have to disclose to your viewers that you used the technology. technology to modify the image. setting.”

This seems about right, although “this could have happened if you just took more photos” is a very vague standard to try to enforce. Furthermore, this standard is in strong tension with YouTube’s decision to have her own Generative AI tools always label synthetic content as such: photos edited with Magic Editor photos have metadata showing that they were edited with AI, but Best Take photos do not.