Thank you all. Let me go ahead and talk a bit about our approach to AI. I want to talk a bit about labels and watermark. First, let me start by saying that we are investing in tools to help users understand when they are interacting with generative AI content. As we invest in more and more capable models, we also care deeply about investing in AI responsibility. This includes having tools to identify synthetically generated content whenever you encounter it. So these days when you look at a synthetic image, it is so impressive how real it looks. We realize that it has a lot of creative potential, but it can also present new risks, like enabling creators to spread false information. So the key question facing all of us is how to strike the right balance between opportunities and risks.

Keyboard shortcuts

j previous speech k next speech