The Google Veo 3 AI video generator now incorporates sound. It might have crossed the uncanny valley

That's impressive. The golden era of mediocrity is upon us, prepare yourselves.
There always was bad content made by people who didn't care, and people who care will still probably keep doing good stuff without using that. The problem is that from now on we will have a lot more people who don't care, now being able to produce things, bad content will flood us, it's going to be worse than ever.
 
I think this is really impressive for how fast the technology is advancing. The question is: how long before we can generate video like this locally?

I can see this being a useful tool in filmmaking, though perhaps not in the way you might imagine. Rather than making an actual film like this, being able to throw a concept together and essentially "demo" a scene as a pitch, or for your actors, or whatever to say "this is my vision," could be an incredibly powerful tool.

Seeing something in your mind might be great, but sharing that vision with someone else can be challenging. Try describing a dream to someone. They're never going to imagine it like you did. So being able to perfectly tweak something like this to your exact specification is a really good way to get that scenario across. AI video, obviously especially as it improves, could have a TON of applications in the future.

In theory it could get advanced enough that it does mimic film or television. Imagine being able to make a show that adapts week to week based on audience input. Silent Hill Ascension actually already tried this but failed because it was poorly written and poorly executed. The IDEA was novel though.
 
Man, AI sure struggles with people's gaze.

Impressive nonetheless, but it always feels like they aren't looking at each other. This stuff is going to be amazing for porn
 
If you know what you're seeing you can spot some imperfections, specially in the direction the eyes are looking, for example, but hot damn this is incredible.

A couple of more years and you won't be able to tell what's AI and what it isn't.
 
Last edited:
At first I was like wow, then I started to notice the issues. You can clearly see people's likeness. The Karate guy is blatantly Carsten Nørgaard.

There's a lot of data it uses that can be found in existing videos somewhere.
 
I could easily recognize A.I generated comments and fake accounts on YouTube or X a year ago. Today? I don't see them much. Recently I got followed by a bunch of "pre debut vtuber accounts" on X and they are so diverse and well made that if it wasn't for the fact that I got these followers suddenly, within few hours, I would never guess they were fake.

They are almost there with videos I see. This sucks. They could fool me with some of them.

eeg49kb0yrsd1.png
 
This is just a treat for those stock video websites, that so many YouTube channels use.

But in games, you need to have a thing called art direction, and consistent designs, and ai can't really do those well.
 
Not too long and we'll get AAA games in months again, rather than the the years its been taking.

But yeah - this will also create a massive problem of employment, which is ironic given tech jobs were sold as future proof only to be amongst the first lost to the what is still only the initial AI wave.

What is happening now in tech is similar to the automation of car manufacture that led to the collapse of major employers and a lot of jobs becoming redundant. Presumably Silicon Valley will resemble 80s Detroit shortly - if it doesn't already with all the remote working.
 
fake as fuck in big parts and actually hard to watch imho, and that`s a curated PR version. No doubt this will continue to make progress, but right now....yuck
 
Last edited:
Much better than the other video in the thread.
There are a few fused people in th background, but yeah. As long as there is little movement in the scenes that model is very close to fooling me.
I bet the fake news channels of every dictatorship in the world are getting sweaty palms right now....

Can you imagine what you can do if you produce AI videos of someone burning a Quran f.e. if you spread it on the right platforms?
We need automated AI-content checks on social media, fast.
 
Last edited:
It's an impressive stunt, but will it ever be more than that? I have a hard time believing it. I put this stuff in the same bucket as a guy beatboxing the Super Mario theme on a flute. Yeah, it's a neat trick, but will it end up on my Spotify Wrapped at the end of the year? Of course not. The prevailing narrative around AI is 'it's crazy what AI can do' and that appears to be the overriding appeal: it's impressive that a machine can mimic parts of the real world so accurately, but that's all I'm getting right now.
 
Looks fake as fuck

It's definitely possible to feel that something's off.

But, remember, AI video just a year or two ago:



I think you'll already see people be fooled by the google video when not in "detective-mode". In the near future it'll likely become near impossible to spot the difference.
 
Last edited:
What? The video looks impressive as fuck. Some sync issues with the audio but considering it's generating voices on the fly.

I said 5 years we'll have full movies being generated, I think I need to wind that down a little.

Everyone else, start retraining your job is about to become redundant.
It's clear it's not perfect. The facial expressions don't match with what's actually happening. The movement doesn't feel totally realistic. The skin looks weird, almost like it's too clean or has been smoothed out too much. I mean shit, it's still impressive, but you can tell it's not real.
 
That's impressive. The golden era of mediocrity is upon us, prepare yourselves.
There always was bad content made by people who didn't care, and people who care will still probably keep doing good stuff without using that. The problem is that from now on we will have a lot more people who don't care, now being able to produce things, bad content will flood us, it's going to be worse than ever.
Upon us? Any more mediocrity then we'll have to start rating content with negative stars.
 
Stop posting AI shit
There are definitely more constructive ways to express concerns about content on a forum than "Stop posting AI shit." Here are some better approaches, depending on the underlying reason for the frustration:


If the AI posts are off-topic:

Better approach: "Hey everyone, I've noticed a lot of posts about AI lately. While it's an interesting topic, it seems a bit off-topic for a forum dedicated to [original forum topic]. Could we try to keep the discussions focused on [original forum topic]?"

Why it's better: This approach is polite, explains the perceived issue, and reminds users of the forum's intended purpose. It also implies that AI discussions might be welcome elsewhere, but not in this specific space.


If there's an overwhelming volume of AI posts:

Better approach: "I'm finding the sheer volume of AI-related posts a bit overwhelming and it's making it hard to find discussions about [other relevant forum topics]. Perhaps we could create a dedicated 'AI Discussion' thread or sub-forum for those who want to talk about it, to help keep the main feed clear for other topics?"

Why it's better: This acknowledges the interest in AI while suggesting a solution for organization. It's collaborative and proposes a way for everyone to get what they want.


If the AI posts are low quality or repetitive:

Better approach: "I've noticed some recent AI posts that seem to be [e.g., repeating the same basic questions, or sharing very general, unoriginal content]. To keep the quality of our discussions high, maybe we could focus on sharing more in-depth analyses, specific case studies, or new breakthroughs in AI?"

Why it's better: This focuses on quality rather than simply shutting down a topic. It provides constructive criticism and guidance on what would be more valuable contributions.


If you simply aren't interested in AI:

Better approach: "Just a thought for the forum admins or anyone else who feels similarly: I'm personally not very interested in AI discussions and they're starting to dominate my feed. Is there a way to filter or categorize posts so users can more easily see the topics they're interested in?"

Why it's better: This frames it as a personal preference and a usability suggestion. It doesn't tell others what to do but asks for tools to manage one's own experience.


General Tips for Constructive Forum Feedback:

  • Be specific: Instead of a vague complaint, explain what the problem is.
  • Suggest solutions: Don't just complain; offer a way to make things better.
  • Use "I" statements: Frame your concern from your perspective ("I'm finding...") rather than making accusatory statements ("You are posting...").
  • Be polite and respectful: Even if you're frustrated, maintaining a civil tone is crucial for productive dialogue.
  • Check forum rules: Before posting, quickly review the forum's guidelines. There might already be a rule or a designated place for feedback.

🤖
 
Last edited:
People complaining that it's not good but their benchmark is actual reality not "good or bad" cgi. That's all you need to know that the goalposts shifted.
 


State of AI video two years ago. Imagine what it'll look like two years from now?
It's so good I expect TV & movie productions start using it for special effects shots.

It's a bit sad the AI got its hands tied, it won't allow to render death, bodily harm and such. So it makes this video unintentionally hilarious when everyone blast away at the walls.

 
Last edited:
Not quite there, but especially cinematics could save much money on actors, motion capturing etc. and time in the future.
 
I will be impressed when they crack comic timing and delivery to actually make one of these jokes even remotely funny.
 
Last edited:
This is going to have a huge impact to the gaming dev world. Now even small devs can produce top quality cut scenes. I wouldn't want to be in the industry that supports gaming cut scenes right now.

honestly, you could of said these were just clips from your average netflix slop and i would of believed you.

to keep OT: Like all AI, there are the bad and the greedy that can be mentioned about this. But i can also see how awesome it could be for directors or producers to try and pitch an idea or show with true vision using something like this. We could potentially see more and better shows being greenlit because the dumb dumb money people and execs with no imagination can see how a story could potentially be visualised.
 
Last edited:
It's an impressive stunt, but will it ever be more than that? I have a hard time believing it. I put this stuff in the same bucket as a guy beatboxing the Super Mario theme on a flute. Yeah, it's a neat trick, but will it end up on my Spotify Wrapped at the end of the year? Of course not. The prevailing narrative around AI is 'it's crazy what AI can do' and that appears to be the overriding appeal: it's impressive that a machine can mimic parts of the real world so accurately, but that's all I'm getting right now.
4 years ago this was all impossible. So yes, it will be more than this. And it will one day be more than just 'impressive'. If we take Veo alone, imagine all those ads you see with random unknown actors. Those all just got a lot cheaper for companies to mak and distribute.
 
It will be a paradise for scammers. My stepfather can no longer recognize that many of them are fakes, and they are mostly just pictures for now. With videos, it only gets worse.
 
Last edited:
We're gonna get to a point where 75% of the population will get AI-scammed into abject poverty and the remaining 25% won't believe anything they see on a screen ever again.
 
And thats supposed to be positive?
It's supposed to be scary and sad for our jobs and future. :(

AI is progressing very quick. Let's just say you probably don't want your kids to go into computer science for a career unless they are truly top of the class and specifically AI focused.

And that's just one field. Video editing, art, data analysis, finance, law, most fields are being impacted.

And yes, Veo is flawed even with version 3, but it looks a lot better vs 2 and it's been less then a year.
 
4 years ago this was all impossible. So yes, it will be more than this. And it will one day be more than just 'impressive'. If we take Veo alone, imagine all those ads you see with random unknown actors. Those all just got a lot cheaper for companies to mak and distribute.
I mean, if the big win is that marketing teams get to make ever more samey and irritating pre-roll trailers for YT, then I guess that's a sort of win, at least in the short term. If you're not employing teams of creatives in the creative side of marketing, you'll end up with a lot of bland ad work, spun up from prompts by some product manager who hasn't a creative bone in their body. The returns will be rapidly diminishing because the cost might go down, but so will the overall efficacy of these campaigns.

For meeting short-term profitability targets, I can definitely see the utility here, but in the long term, you're replacing innovation with iteration and speeding up the brain drain. When you've finally managed to replace all your creative teams with a giant, cheap AI that turns stuff around in minutes rather than weeks, you'll have nothing left but a bunch of middle-aged, middle-class, middle-managers feeding it dull prompts and getting the same iterative dross out the other end. And because it's cheap and easily, they'll make dozens of cheap crap ones, rather spend three or four months putting together something great. The internet, meanwhile, is flooded with all this bland AI content created from the same training data sources, which in turn becomes the training data itself - a sort of information inbreeding that makes everything gradually more dull and unremarkable.

Diminishing returns force product managers to panic and produce even more variants and the situation worsens...

The thing is that we've already seen what happens when algorithms make content, even with human steer: it's bland. Big publishers have been running business on the basis of big data for the last fifteen years and you can already see what that's done: more and more iterative dross informed by business intelligence and data analytics that no matter how pretty or impressive, is repetitive, iterative and fails to draw in audiences. The MCU and DCEU both fell prey to this kind of cultural inbreeding, where brands are creating content that's only ever in conversation with itself based on 'proven' templates, and output is becoming noticeably more malformed and wall-eyed with every new iteration.

Games, in broad brush strokes, are doing the same, relying on an iterative process to force a trend line from 2018 to keep going up and up into infinity, Disco Stu style. It's not working and the games that are blowing up now are the ones that ignore the algorithms and do something that seems right to them: Larian bring back the turn-based CRPG junked by most publishers 20 years ago and create one of the biggest-selling games of the generation. Balatro sells a bajillion copies and scoops up awards for a pixel-art poker roguelike. Clair Obscur mixes old-school FF with Persona and Sekiro and sets it in a fucked up Paris and create the year's most talked-about title. Disco Elysium revives the old-skool point-and-click adventure to create an off-the-wall political drama about an alcoholic cop, sells gangbusters. Helldivers 2 is a tongue-in-cheek send-up of corporate democracy, riffing hard on the work of Paul Verhoven that belongs in some genre all of its own and is Sony's fastest-selling title to date.

Generative AI just doubles down and turbo charges that already failed iterative model and speeds up the much need collapse, turning out twice much of the same shit we're already ignoring at a fraction of the time and they'll sell even less. The real winners in the long-term will be the creatives and the people who enjoy great ideas. The big losers will be shareholders and those bizarre fan boys who flag-wave for mega corps.
 
Last edited:
As a post-producer and videomaker, what impress me most about AI is how close to reality everything starts to looks. As CGI is becoming worse and worse in blockbuster movies (see Marvel movies or the last Final Destination), we have AI that is starting to show things indistinguishable from reality at basically zero budget.

In some years the movie industry will be completely revolutionised.
 
Last edited:
I keep thinking about Jensens nightmarish CES presentation last year. AI robot arms cooking like a chef. AI HR agents. AI cameras watching warehouse workers and highlighting people standing still.
 
What's the point of this?
Basically where I'm at. Something I can glance at and say, "neat", but that's about it.

I think AI IS capable of some interesting things. But I'm not really sure what it's accomplishing with stuff like this. Especially the weird following that some out there have with wanting it to replace movies, etc. Which is something I don't understand at all.
 
I mean, if the big win is that marketing teams get to make ever more samey and irritating pre-roll trailers for YT, then I guess that's a sort of win, at least in the short term. If you're not employing teams of creatives in the creative side of marketing, you'll end up with a lot of bland ad work, spun up from prompts by some product manager who hasn't a creative bone in their body. The returns will be rapidly diminishing because the cost might go down, but so will the overall efficacy of these campaigns.

For meeting short-term profitability targets, I can definitely see the utility here, but in the long term, you're replacing innovation with iteration and speeding up the brain drain. When you've finally managed to replace all your creative teams with a giant, cheap AI that turns stuff around in minutes rather than weeks, you'll have nothing left but a bunch of middle-aged, middle-class, middle-managers feeding it dull prompts and getting the same iterative dross out the other end. And because it's cheap and easily, they'll make dozens of cheap crap ones, rather spend three or four months putting together something great. The internet, meanwhile, is flooded with all this bland AI content created from the same training data sources, which in turn becomes the training data itself - a sort of information inbreeding that makes everything gradually more dull and unremarkable.

Diminishing returns force product managers to panic and produce even more variants and the situation worsens...

The thing is that we've already seen what happens when algorithms make content, even with human steer: it's bland. Big publishers have been running business on the basis of big data for the last fifteen years and you can already see what that's done: more and more iterative dross informed by business intelligence and data analytics that no matter how pretty or impressive, is repetitive, iterative and fails to draw in audiences. The MCU and DCEU both fell prey to this kind of cultural inbreeding, where brands are creating content that's only ever in conversation with itself based on 'proven' templates, and output is becoming noticeably more malformed and wall-eyed with every new iteration.

Games, in broad brush strokes, are doing the same, relying on an iterative process to force a trend line from 2018 to keep going up and up into infinity, Disco Stu style. It's not working and the games that are blowing up now are the ones that ignore the algorithms and do something that seems right to them: Larian bring back the turn-based CRPG junked by most publishers 20 years ago and create one of the biggest-selling games of the generation. Balatro sells a bajillion copies and scoops up awards for a pixel-art poker roguelike. Clair Obscur mixes old-school FF with Persona and Sekiro and sets it in a fucked up Paris and create the year's most talked-about title. Disco Elysium revives the old-skool point-and-click adventure to create an off-the-wall political drama about an alcoholic cop, sells gangbusters. Helldivers 2 is a tongue-in-cheek send-up of corporate democracy, riffing hard on the work of Paul Verhoven that belongs in some genre all of its own and is Sony's fastest-selling title to date.

Generative AI just doubles down and turbo charges that already failed iterative model and speeds up the much need collapse, turning out twice much of the same shit we're already ignoring at a fraction of the time and they'll sell even less. The real winners in the long-term will be the creatives and the people who enjoy great ideas. The big losers will be shareholders and those bizarre fan boys who flag-wave for mega corps.
I completely agree.
 
They (will be) tek yer jobs!
But wouldn't it be fascinating if the majority finally found out we can automate most of our work life and actually don't need it all?
Hahahaaa.
...
 
Imagine injecting yourself into AI and producing a reenactment of your life. Fine tune all the fake corny stuff, but that would be an interesting gift to hand someone in your will. "Here's a 3 season TV of the memories you shared with your loved one". The possibilities are insane or perhaps inject yourself into a movie or adult film. It can't replace human creativity because a lot of this stuff isn't fine tuned, edited, and feels soulless. I think it's a lot like the smart phone. For a while we had flip phones and basic Nokia screens. Now we have phones with HD cameras that can run full fledge games basically. It's eventual. Perhaps the industry will be a lot more creative with AI than VR. I feel like VR takes a nose dive because it's needing revisions, games or apps made for it, and it gets pricey if you want a better experience.
 
Top Bottom