As far as the AI & 3D modeling, or deepfake videos, before buying into the panic about the topic, we need to think about the limits of this technology.
To make convincing deep fake videos requires some serious computing power, as in, any computer you might have that COULD handle the algorithm (ridiculous amounts of RAM needed) would require weeks of training just to get the algorithm to learn how to create the modeled face.
Then you would have to render the video, sure you can lip synch whatever you want the person to say, but unless you can ALSO convincingly fake their voice, nobody would believe the clip of Obama saying what you wanted if the voice is one of the chipmunks. (Yes, there also exists similar algorithms to recreate a person's voice, but again, you're back to training, though training audio is far less memory intensive than video)
And sure, there are algorithms that are far lighter to fake a video, but those algorithms could never produce anything that would be convincing.
Bottom line; for the type of panic claimed about this tech, anything convincing is going to be a high level production. Seems much better as a cover story to deflect when real videos are released.
As far as the AI & 3D modeling, or deepfake videos, before buying into the panic about the topic, we need to think about the limits of this technology.
To make convincing deep fake videos requires some serious computing power, as in, any computer you might have that COULD handle the algorithm (ridiculous amounts of RAM needed) would require weeks of training just to get the algorithm to learn how to create the modeled face.
Then you would have to render the video, sure you can lip synch whatever you want the person to say, but unless you can ALSO convincingly fake their voice, nobody would believe the clip of Obama saying what you wanted if the voice is one of the chipmunks. (Yes, there also exists similar algorithms to recreate a person's voice, but again, you're back to training, though training audio is far less memory intensive than video)
And sure, there are algorithms that are far lighter to fake a video, but those algorithms could never produce anything that would be convincing.
Bottom line; for the type of panic claimed about this tech, anything convincing is going to be a high level production. Seems much better as a cover story to deflect when real videos are released.
two words
vidia gams
once you have a model, and texture map, its only the bugs that ruin experience. hell halo masterchief collection has upped the ante.