synnbiob
Related Articles
I had to read the news about six times before it fully sunk in: Meta, the owner of Instagram and Facebook, is experimenting with integrating custom AI content into our Instagram and Facebook feeds. To put it another way, Meta is exploring the idea of putting more recommended content (from accounts we don’t follow) into our feeds, and it would be entirely generated by Meta’s AI.
As the person who reviews AI image generators for CNET, including Meta AI, let me tell you: This is a terrible idea.
These photos, made with its AI assistant Meta AI, will be “based on your interests” and could potentially “feature you, so you can be the star of your own story and share your favorites with friends.” So an AI version of your face could randomly pop up in your Instagram feed. Meta, respectfully… what? Why?
Meta dropped a lot of news about its virtual assistant Meta AI this week at its annual AR/VR event Meta Connect, including a new voice featuring a variety of celebrity voices, better video dubbing translations and new visual upgrades that let it “see” your photos and edit them. But this news, buried near the end of its press release and omitted entirely from the keynote address, is important, and we need to talk about it. Mainly, we need to beg Meta not to let this experiment result in a real update.
As AI content generation services get better, it’s getting harder to escape the deluge of AI content (aka slop). It’s also getting harder to tell real photos apart from AI ones. I admit, I was pretty impressed with Meta AI’s Imagine feature, which acts as an AI image generator embedded in the chatbot. But that doesn’t mean I want it to clog up my feed with weird images of me as an astronaut or whatever.
Earlier this year, Meta started labeling content that was AI-generated, which was a great idea. But it wasn’t a perfect tool, and photographers started getting slapped with AI labels for images that were edited, not AI-generated. The new recommended posts would have a similar label that says “Imagined for you.” But labels are too easy to overlook, which is why I’m still concerned that rolling out this update would further blur the line between what’s real and what’s AI. And that’s really the last thing we need, when mis- and disinformation is easier to create convincingly than ever before, thanks to AI.
This is an example of what these AI-generated recommended posts could look like.
On the privacy front, CNET reached out to Meta to confirm whether you can remove AI-generated posts from your feed and if we can expect to see AI posts start to pop up while it’s in the testing phase, but we didn’t receive a response by the time of publication. We do know that US users cannot opt out of having publicly shared posts used for training its AI models.
Mark Zuckerberg told The Verge that adding AI content to social feeds is the next “logical jump” for the social platforms. But I really can’t imagine an AI image that’s going to be so funny, so relevant or otherwise worth my time that it’s going to make up for all the potential problems it could produce. Meta has spent so much time convincing us that if we want to use Meta AI, we can find it everywhere — even if we didn’t want it there, like in Search. Why do we need it in our feeds, too?
My Instagram feed is already a mess, mostly clogged up by random accounts I don’t follow, recommending content I might enjoy. I have to click through at least three ads to get to one Story from someone I follow. And I’ve put up with it, mostly because it’s still my main online social hub. But the overall user experience of scrolling Instagram and Facebook has gone downhill in the past 10 years. We’ve long since moved past its original purpose to connect with friends and family, much to some users’ disappointment, including me. Adding AI-generated recommended content to the mix feels like it will be the final nail in the coffin of my Instagram feed.
