I'll give it a few decades max until deep fakes in reality are a thing. Perhaps everyone walks around with smart glasses or contact lenses that project an AR layer on top of reality. When you hack someone's device, you might meet them and as long as they don't take or turn off their device and they won't grab your face with their hands to feel if it's real, you could inject a deep faked face in their device and you can look like their boss, like their spouse, like the president, like anyone. For the case, anyone from Netflix is reading this, yes you can have it for a Black Mirror episode 😄 Could be a sequel to an episode where contact lenses recorded your whole life with every detail.
It was already scary, and as someone who is always up to date with the latest AI news i should already be used to it but even still feels a bit too scary this time
News & political deep fakes are the scariest part about it. It was hard enough to trust the news before AI and deep fakes came along, and now it's about to get even more dystopian.
You don't have enough of an imagination for nefarious use apparently. Have a project idea to make you money in spite of the risk of misuse to humanity per chance? Just a wild guess.
I think CG avatar is really interesting for someone who would want to create a semi-anonymous channel on youtube or similar platform. You can still use your real voice, speak as yourself, the viewer sees an engaging character, but the content creator gets to keep their appearance secret.
Can alter your voice without it being the obvious ai bot reading script. The channel there I ruined it showed a bit of how they alter voice to mix artists to other songs
I don't know when, but I am certain that in the future video games will be rendered by AI in real time. I still think that humans remain necessary to create a coherent experience, so to speak, composing the arrangement. But it will be very different to what we have today. Both, unsettling and exciting outlook, considering what else might change along the way.
Sadly, Deep Live Cam uses inswapper, which is like 3 year old technology at this point (and very blurry, 128x128 pixels for the whole face). It is nothing new. Just a new interface for it.
bored of the "this is scary", "is this too much" mindset. same bullshit was said with SD1.5. No, its just new and not every new thing is society breaking. Lets stop giving the doomers food.We will adapt in a matter of weeks and we will become soo aware that we will be able to quickly pick out the nuanced differences in no time. Now...I look forward to testing this with monster faces. dungeons and dragons type content would be very cool. especially zoom type stuff.
I agree. The conversation should be centered around creativity. Doom scenarios can give politicians fuel to use fear tactics to pass laws. This could ban the talented AI research by individuals and make large corporations the gatekeepers.
I wonder how much time will take until politicians, who promised this and that before elections, will say it was the AI, not them, that made those promises
They don't make this promises sitting in front of their pc though, but in congress, official channels or in front of multiple journalists. Still we already went too far.
Hello Olivio, what artist is it with the name 'Paleopterix'? You gave the painted image a title: 'Abstract shapes and colors to express strong emotions'. I can't find it via google,.Can you give me a hint? I like the painting too!
I could see the lip synching feature being really handy to localize and redub foreign language films. On the other hand, catfishing is going to be taken to a whole 'nother level now. You could quite possibly be tricked into thinking you're talking with somebody who you know in real life ("help, I'm in an emergency and I need to to Venmo me some money right away!") but it's really some scammer in another country.
Would have zero effect due to fact that everyone would put the fakes out there so the only thing people would believe is televised broadcast. It like all the spam email for you winning something.
All I can say is that I've been working for a few years now to integrate this tech into my filmmaking and I cannot wait to show the world with what I've come up with.
It is times like this I wish MS had not killed of the Kinect sensor, it was ahead of it's time and would have been perfect for this kind of motion capture. Video AI is approaching uncanny value, but lack the niceties of what you would expect from a scene, specular hights, diffraction, details like skin texture and so no. Overall it is getting close but it is the smaller details that give it away for now.
@@flickwtchr It some case we're already there for those with the abilities, all this is doing is democratising the tech so anyone can do it. Which is why there is such a demand for forensic tools to analyse images.
The CG Avatar part seem like a complete scam to me... If you follow the messages, first he says: "Since there were many questions, I will write this generation method in the reply section" Like, what do you even mean many questions and how does putting the generation method in the reply section help?? Then we scroll down to the reply section as he said he's gonna put the method in there, and says: "We will tell you how to generate later, if interested, please register with the link below" And guess what, it's a QR code.......!
Deep fake? Boy we're in trouble...remember all those people who after CONVID all rushed to McDonalds? We're screwed big time and that enourmous jousting robot is the future....💩
UPDATE: It looks like "Deep Live Cam" is a fork of another roop project. Thank you everyone for letting me know!
I'd still like to try it (or the source project) and dial into the next team meeting at work as Duke Nukem.
Another excellent update! Thank you, Olivio😊
RIP Online dating.
Trust nobody by audio or video.
@@FusionDeveloper for real LOL
Online dating is dead for almost a decade anyway... As well as a dating in general. So, not such a big loss.
Mind blown! -- and rather frightening (a flesh handshake will soon be the new standard for "this is me.")
Logan's Circuit? Me, first
Flesh Handshake 😛
I'll give it a few decades max until deep fakes in reality are a thing. Perhaps everyone walks around with smart glasses or contact lenses that project an AR layer on top of reality. When you hack someone's device, you might meet them and as long as they don't take or turn off their device and they won't grab your face with their hands to feel if it's real, you could inject a deep faked face in their device and you can look like their boss, like their spouse, like the president, like anyone.
For the case, anyone from Netflix is reading this, yes you can have it for a Black Mirror episode 😄 Could be a sequel to an episode where contact lenses recorded your whole life with every detail.
Thanks Olivio for the great info, as always. I'm glad you brought back the blue cat at the end... 😀
9:57 googley eyes 👀
It was already scary, and as someone who is always up to date with the latest AI news i should already be used to it but even still feels a bit too scary this time
News & political deep fakes are the scariest part about it. It was hard enough to trust the news before AI and deep fakes came along, and now it's about to get even more dystopian.
This is what I"m waiting for! Love it, It's not too much! ♥♥♥ Thank you for sharing this! So Excited!
You don't have enough of an imagination for nefarious use apparently. Have a project idea to make you money in spite of the risk of misuse to humanity per chance? Just a wild guess.
I gotta go through all that, Thanks for another amazing Video Olivio
Thank you so much for sharing! Big fan of your videos 👏
The Flux vids are image2video with Kling, not Flux Video AI. Flux hasnt released its video generation models yet.
Yeah! The end screen came back! 💪🏻🔥
Except from a link on twitter with a video, is there any way for us to try CG avatar...
I think CG avatar is really interesting for someone who would want to create a semi-anonymous channel on youtube or similar platform. You can still use your real voice, speak as yourself, the viewer sees an engaging character, but the content creator gets to keep their appearance secret.
Hopefully this will be banned by youtube that has already become a cesspool full of AI generated garbage.
We had PNGtubers, Vtubers, and now we can have AItubers. I'm down.
Can alter your voice without it being the obvious ai bot reading script. The channel there I ruined it showed a bit of how they alter voice to mix artists to other songs
the langauge understanding of these models are phenomenal.
I don't know when, but I am certain that in the future video games will be rendered by AI in real time. I still think that humans remain necessary to create a coherent experience, so to speak, composing the arrangement. But it will be very different to what we have today.
Both, unsettling and exciting outlook, considering what else might change along the way.
Sadly, Deep Live Cam uses inswapper, which is like 3 year old technology at this point (and very blurry, 128x128 pixels for the whole face). It is nothing new. Just a new interface for it.
Out of interest, do you know if there is something newer out?
@@finnishnutmilk FaceFusion
@@henryruhs Facefusion is the undisputed king. The guys behind that one are geniuses. 😉
Of the tests I've seen, the output seems better than facefusion
@@henryruhs FaceFusion is just inswapper_128 yet again. There is nothing newer out.
bored of the "this is scary", "is this too much" mindset. same bullshit was said with SD1.5. No, its just new and not every new thing is society breaking. Lets stop giving the doomers food.We will adapt in a matter of weeks and we will become soo aware that we will be able to quickly pick out the nuanced differences in no time. Now...I look forward to testing this with monster faces. dungeons and dragons type content would be very cool. especially zoom type stuff.
I agree. The conversation should be centered around creativity. Doom scenarios can give politicians fuel to use fear tactics to pass laws. This could ban the talented AI research by individuals and make large corporations the gatekeepers.
"more people can express themselves." 您的這句話真是非常畫龍點睛。在技術創新之下,會造成混亂,但是真正的意義如閣下所言,是助長了知識與新技術的傳播,也鼓勵更多人去分享彼此的看法與思想。思想的傳遞與交流是真正最重要的。
this is going to be a nightmare with scammers.
Then maybe finally something will be done to rid the world of scammers. Currently they're allowed to roam free.
I want to watch my favorite youtubers speaking my language
Can you post a link to CG Avatar AI please? The one that you provided is taking to X and showing same example and link to their website.
I'm not sure if the CG avatar one is something new or if it's LivePortrait
Probably LivePortrait. I tried following the thread, but having to click on each reply to translate from Japanese was too tedious.
I wonder how much time will take until politicians, who promised this and that before elections, will say it was the AI, not them, that made those promises
They don't make this promises sitting in front of their pc though, but in congress, official channels or in front of multiple journalists. Still we already went too far.
Fair number of "journalists" ignore or even defend the politicians though.
@@justintime3384
Interesting. Would probably take a lot of work to do what I’m thinking but this is a major leap forward
Hello Olivio, what artist is it with the name 'Paleopterix'? You gave the painted image a title: 'Abstract shapes and colors to express strong emotions'. I can't find it via google,.Can you give me a hint? I like the painting too!
Flux is not a video model. you should really try it out
Yeah some guy on X saying so.. nothing there at all.
Yup. Flux/Midjourney to create the starting image, then Kling to generate the video.
They will at some point since a demo was present on their website
the smoke sim will reduce in most cases the sim time to zero, amazing
Both images are Thor: old Norse vs modern Americanized Thor
Thank you Olivio🎉
The one bloke still had 6 fingers though….🤦♂️
awesome news runwayml really improved
Happy Saturday! 🥳
Yeh that Picasso style was really cool...
I tried Deep Live Cam and it is absolutely not working, not live nor non real time. only got flickering faces
it's absolutely working. did you select the proper --execution-provider ?
@@GigaTele used cuda, which should be the right one to run it on my GPU
"It has the potential to bring information to the people ... to liberate ..." ok
I could see the lip synching feature being really handy to localize and redub foreign language films.
On the other hand, catfishing is going to be taken to a whole 'nother level now. You could quite possibly be tricked into thinking you're talking with somebody who you know in real life ("help, I'm in an emergency and I need to to Venmo me some money right away!") but it's really some scammer in another country.
9:56 look at his eyes !! 🤣
Much much BIGGER potential (scientist eyes)
Tell me you want to interfere with American politics without telling me
ReSyncer:
Would have zero effect due to fact that everyone would put the fakes out there so the only thing people would believe is televised broadcast. It like all the spam email for you winning something.
DeepfaceLive has been doing this for like 4 years haha
Romance scammers be like
All I can say is that I've been working for a few years now to integrate this tech into my filmmaking and I cannot wait to show the world with what I've come up with.
wow
warning of Deep-Live-Cam IDP generic troyan
because he is a robot... did it for me 😂😂😂😂😂
I feel like I need to make a safe word with my loved ones now to keep from being deep faked lol
"Flux video" is actually just using Flux images inputted into Kling, so it's really just Kling video
It is times like this I wish MS had not killed of the Kinect sensor, it was ahead of it's time and would have been perfect for this kind of motion capture. Video AI is approaching uncanny value, but lack the niceties of what you would expect from a scene, specular hights, diffraction, details like skin texture and so no. Overall it is getting close but it is the smaller details that give it away for now.
But having absolutely no ability to discern what is real and what is fake will just be so great, right?
@@flickwtchr It some case we're already there for those with the abilities, all this is doing is democratising the tech so anyone can do it. Which is why there is such a demand for forensic tools to analyse images.
I listen to you on Russian with sync translation provided by Yandex browser. I would say that lips sync not so nessessory as better translation
the X links says "Workflow & Prompts below:" but there is nothing.
The film industry has been working with this technology for years, but I fear that it will be banned for us little users.
Time to use verified security certificates to digitally sign personal videos
Facefusion is better than everything else
E-fact 😄
The CG Avatar part seem like a complete scam to me... If you follow the messages, first he says:
"Since there were many questions, I will write this generation method in the reply section"
Like, what do you even mean many questions and how does putting the generation method in the reply section help??
Then we scroll down to the reply section as he said he's gonna put the method in there, and says:
"We will tell you how to generate later, if interested, please register with the link below"
And guess what, it's a QR code.......!
How to one gain access to the CG Avatar one?
same question
nice!
The future of CZcams is AiTubers, which are indistinguishable from the real thing that streams 24 hours a day and does their own research.
I think CG Avatar is just a demo, not something you can actually use.
🙃😉😊
Raising the prices on my furry content as soon as this is revised more
guys just have to get their lazy butts out in real life...no AI in person
How about a cartoon?
Deep fake? Boy we're in trouble...remember all those people who after CONVID all rushed to McDonalds? We're screwed big time and that enourmous jousting robot is the future....💩
So your mind is blown by these tools but you don’t even bother to install them and try yourself? Just showing github frontpage and a few tweets?
iT'S NEVER TOO MUCH FROM AI !!! We want more from AI!! more advanced stuff from AI!!