Google Veo 3 (AI video generation) is insane

I am an AI doomer. Not that I think ASI will neccessarily kill us all (though it's very possible and big tech is steaming towards it with no plan to deal with it that stands up to logical scrutiny), but on the near-future, non-AGI/ASI harms - like endless genAI slop replacing traditional media. Social media was/is bad enough when Zuck turned it into an addictive slop dystopia, but it's going to be absolutely supercharged by genAI (and he has recently admitted that's his goal in a rare mask-off moment in an interview). People forgetting how to code and do other knowledge work because AI does it for them. Inability to distinguish genAI from reality will cause all manner of big problems.

The next few years are going to be crazy and normies (and economists who price in 0.5% GDP growth from AI) haven't a clue what's coming. Case in point, I showed a Sora video (the woman walking in Tokyo) to a fairly intelligent 20 year old relative when it launched and she didn't think it was a big deal at all. Whereas I was chilled to the bone in the way that I expect a medieval peasant would have been if you could go back in time and show them an Airbus A380.
 
The singularity was always said to be around the year 2035, the point when AI gets so good that it'[s impossible to know if what you are seeing, speaking to or hearing is created by humans or a machine.

Given where we are right now, that will happen years sooner and gives further ammo to the whole simulation theory.
 
I've seen an AD and if it's working as they claim, O2 have an AI that's tackling scammers by wasting their time.
 
Wound my daughter up by taking her FB profile pic and using Veo to make her twerk:

Me: "so this is what you get up to in Uni!!!!"
Daughter: "WTH?! I swear i didn't do that!"

It looked very real - scarily real :(

My son appreciated being a Jedi though :)
 
Last edited:
lovelyhead said:
This makes me feel bad.

I don't think this is a positive for humankind. It won't be long before AI video is undistinguishable from real video. At that point YouTube is dead. Movies are dead. Music is dead. Podcast's are dead. You can never trust any video or audio ever again. Millions of people out of work. It will get to the stage where AI has created the bulk of the internet and AI is using AI generated data to train itself.

I am no expert on AI so the above might not be accurate but those are my immediate thoughts. It just seems like this is going to be a huge net negative for humanity.
Click to expand...

I don't agree with you at all. You're not going to watch an AI video of someone building a PC or fixing it a car or playing guitar - it would be way too hard to make that stuff and make it genuinely interesting.
 
dirtychinchilla said:
I don't agree with you at all. You're not going to watch an AI video of someone building a PC or fixing it a car or playing guitar - it would be way too hard to make that stuff and make it genuinely interesting.
Click to expand...

You don't think AI will get good enough where you can prompt it to make a video about building a PC with such and such components and use Linus Sebastian as the host? It was only a couple of years ago that will smith was eating spaghetti.
 
dirtychinchilla said:
Not with a genuinely interesting and unpredictable story line, no.
Click to expand...

I hope you are right but the tech is already advanced enough to make stand up comedy scenes with jokes that made me laugh.

I bet it is already good enough to make a Linus tech tips clip and it be good enough for most people to not be able to tell it isn't ai. The person making the video would just have to think up some banter to prompt the ai with.
 
mrk said:
The singularity was always said to be around the year 2035, the point when AI gets so good that it'[s impossible to know if what you are seeing, speaking to or hearing is created by humans or a machine.

Given where we are right now, that will happen years sooner and gives further ammo to the whole simulation theory.
Click to expand...

Raymond Kurzweil in his 2005 book 'The singularity is near' put it at 2045. Right now, I'd say that's conservative - I think you're right, probably sooner. Not a bad prediction from 2005 though.

Or who knows, maybe it's already here, we just don't know it yet.
 
ivrytwr3 said:
Wound my daughter up by taking her FB profile pic and using Veo to make her twerk:

Me: "so this is what you get up to in Uni!!!!"
Daughter: "WTH?! I swear i didn't do that!"

It looked very real - scarily real :(

My son appreciated being a Jedi though :)
Click to expand...


Hmm well I can guess what other people are using it for if you can just grab pics of social media and get people doing anything...

Not going to end well...time to get off social media I reckon....
 
Bobmunkhouse said:
Hmm well I can guess what other people are using it for if you can just grab pics of social media and get people doing anything...

Not going to end well...time to get off social media I reckon....
Click to expand...
This.

But unfortunately that ship sailed long ago, long before Veo 3 was even a thing (It's just easier).

I try to suggest to friends and family to remain anon unless if no other choice.
 
lovelyhead said:
Time will tell. I reckon within the next 5 years a big name youtuber will put out a video made entirely with ai without telling anyone and it will trick most people.
Click to expand...

Bit of a different way around but I've noticed a increase lately of videos on YouTube covering topics with a youngish person who superficially seems to be talking quite earnestly on a subject seemingly in depth and with well constructed arguments building up the points but after awhile you realise they are talking complete garbage full of fallacious takes and is almost certainly ChatGPT generated content with a ChatGPT script. But quite dangerous as people could easily be convinced by the manner of delivery that what they were saying made any kind of actual sense.
 
My fiancé was asking me last night if I'd seen the new completely transparent phone that Nokia are bringing out...

Had to break it to her gently that it was completely fake, and then showed her the car show video linked above in post #15 to demonstrate how good some of the AI generated stuff is these days. She didn't believe me that it was 100% AI generated!

You're absolutely right that "normies" (i.e. non-tech people) don't have a clue what's possible now, and are likely to believe whatever they see... it's going to be an interesting few years, it reminds me of a police/murder mystery drama/thriller we watched a few years ago - "The Capture":

a guy was charged with murder based on clear CCTV evidence of him doing it, except in the background there was a shadowy government agency who had edited the CCTV footage. Problem is, these days you don't need several hundred £k worth of computers and a team of digital video artists, just 10 minutes of tweaking an AI prompt!
 
I don't really see how any person in any industry where computer software is used to do the job is going to survive...(Apart from developing the AI itself....although AI might get better at doing that too....)

Best start getting into doing something practical/real, because vast swathes of every sector will be replaced by AI.
 
Last edited:
Jono8 said:
I don't really see how any person in any industry where computer software is used to do the job is going to survive...(Apart from developing the AI itself....although AI might get better at doing that too....)

Best start getting into doing something practical/real, because vast swathes of every sector will be replaced by AI.
Click to expand...
Any work where it involves practical (plumbing, food preperation, military etc) or secrecy (Gov, Agencies, etc) or religion/philosophical will likely be essentially safe from AI, since by the nature of those works - AI can't perform the work, be trusted with the information or AI can't replace it ,(eg there's probably stuff only the Pope gets to know about, etc, so another thing on secrecy).
 
Last edited:
Meddling-Monk said:
Any work where it involves practical (plumbing, food preperation, military etc) or secrecy (Gov, Agencies, etc) or religion/philosophical will likely be essentially safe from AI, since by the nature of those works - AI can't perform the work, be trusted with the information or AI can't replace it ,(eg there's probably stuff only the Pope gets to know about, etc, so another thing on secrecy).
Click to expand...

Yeh, it seems like only hospitality, tourism/events and construction/trades and are truly safe from AI.

...but then the competition for those jobs will go through the roof if there is nothing else
 
You must log in or register to reply here.
Back
Top Bottom