It’s hard to tell where anything is coming from these days when it comes to photos and videos.
That’s because you can generate some pretty convincing stuff with nothing more than a text prompt.
And it’s only going to get more impressive from here.
Luckily, a few prescient people out there have taken note and are tackling the issue of media provenance with earnestness but the conversation mainly centers around content created out of thin air.
There’s also the issue of AI editing of images and videos. While not creating an entire work on its own, this area of generative AI also needs attention for the same reasons that people take issues with filters and such on social media: People need to know that what they are looking at is, in some way, not real.
Google is bringing its digital watermark technology to images that are not only generated entirely via AI but also those that use it for editing, the company’s blog reports.
“Starting this week, Google Photos will begin using SynthID (a technology that embeds an imperceptible, digital watermark directly into AI-generated images, audio, text or video) to mark images edited with generative AI using Reimagine in Magic Editor. You may have already seen SynthID used to watermark images created entirely by AI — like those made by Google’s text-to-image model, Imagen. This helps people identify AI-generated content quickly and easily.”
Interestingly, it won’t be able to detect any change, no matter how minute – it has to be something somewhat significant. The company uses the example of changing the color of a small flower in the background of an image as something that might escape detection. We expect that will probably change over time as well with any changes whatsoever being detectable but we will wait and see on that.
Any thoughts that you might have on the future of photography and videography in the age of AI are welcome in the comments.
Be sure to check out our other articles at this link.
