Google’s Veo 3 Has People Crashing Out Over AI Slop

Depending on who you ask, generative AI is either a thrilling tech revolution or an existential threat, and there’s little in-between. It’s hard to blame anyone for an extreme reaction, too, given the magnitude of capital investment, hyperbolic marketing, and rapid progress of generative AI in such a short amount of time. But it’s not just the economics and technical feats of AI that have people losing their minds; there’s also something more philosophical percolating, and it’s driving some people “to the brink.”
People are literally having a mental breakdown over Veo-3 pic.twitter.com/ym5oZDYZGr
— Chubby♨️ (@kimmonismus) May 27, 2025
The latest AI advancement to send people down an existential rabbit hole comes courtesy of Google, which just announced its latest video generation model called Veo 3. As I’ve reported a few times now, Veo 3 is already getting into some wild stuff—turning up the dial on AI slop, deepfaking smooth-brained YouTube content, and potentially upending game development, to name a few things. As it turns out, people are taking note of all of those feats, and some of them are not exactly happy about what they see.
As evidenced by a thread from the subreddit r/artificialintelligence posted this week titled “VEO3 is kind of bringing me to a mental brink. What are we even doing anymore?” Google’s Veo 3 and the implications therein have some people spiraling. “I’m just kind of speechless. The concept of existential crisis has taken a whole new form. I was unhappy with my life just now but thought I can turn it around, but if I turn it around, what is left of our world in 2 decades?” the post’s author writes.
“Actors as a concept are gone? Manually creating music? Wallpapers? Game assets? Believing comments on the internet are from real people? AI edited photos are just as real as the original samples? Voicenotes can be perfectly faked?… Literally what value is being left for us?”
Reactions to the thread are mixed, with suggestions that the author should go “touch grass” or maybe “go to therapy,” but there’s also a chorus in agreement. The consensus from the latter group? AI slop is coming to ruin your art, and there’s not much we can do about it.
I, for what it’s worth, fall unhelpfully in between the two camps. I think there is a deluge of AI slop incoming, and, if we’re being honest, we’re already up to our ankles. Between Veo and OpenAI’s Sora and the clear interest in automating human creativity, I think we can reasonably buckle in and expect the world of movies, music, and entertainment writ large to get a little choppy. Whether any of those efforts to automate entertainment will stick is less obvious. The thing about art is that the kind that people tend to like is the kind that has something substantial to say. Right now, for all of its mimicry, generative AI doesn’t actually have anything to say, because technically all it can do is remix and repeat.
I did more tests with Google's #Veo3. Imagine if AI characters became aware they were living in a simulation! pic.twitter.com/nhbrNQMtqv
— Hashem Al-Ghaili (@HashemGhaili) May 21, 2025
Call me an optimist, but most people can likely sniff out the difference between slop and art, and as much as studios would love to wave a magic wand and rid themselves of human creatives and the cost of their labor, deep down they know that they’d have to Ctrl+Z that move just as fast. That’s not to say there won’t be casualties in the AI age—if there’s one lesson we can learn from mass waves of automation in years past, it’s that labor forces are usually the most affected.
But when it comes to art, things aren’t so simple. Art, at least the good kind, is about human connection, and until AI can think and feel like we do, there’s nothing that can replace that. So, before you crash out over AI slop, just remember: AI still thinks putting glue on your pizza is a good idea, so we may have a few more good years left in the tank.
gizmodo