This website uses AI to generate a fake human face every two seconds

Shawn Knight

TechSpot Staff
Staff member

Deepfakes, or the act of using artificial intelligence and machine learning to plaster celebrity faces onto the bodies of other people, is a hot-button issue in the tech community. Perhaps even more alarming is a new website that demonstrates an algorithm’s ability to create portraits on command of people that don’t actually exist.

ThisPersonDoesNotExist.com, the brainchild of Uber software engineer Phillip Wang, uses generative adversarial network (GAN) technology from Nvidia called StyleGAN to create a new facial image from scratch based on a 512 dimensional vector.

The results, more often than not, are indistinguishable from authentic images. Occasionally, you’ll come across a generated image with some obvious artifacts and if you take extra time to analyze the offerings as a whole, you may notice that teeth sometimes look a bit unusual with random discoloration or spotting.

Still, for technology that’s only in its infancy, it’s darn impressive… or outright scary, depending on your take. If it’s this “easy” to generate a fake face, it won’t be long before we are able to create other fake things... evidence, environments, societies. Maybe Elon Musk’s theory about us being in a simulation isn’t all that far-fetched?

Permalink to story.

 

Misagt

TS Evangelist
We are quickly reaching the point were we won't be able to trust video anymore, are at that this point already for photos. It's going to be an interesting time once realistic video and audio of a person can be created of them doing anything.
 

Misagt

TS Evangelist
If anyone should be scared of this, it's actors soon you all will be out of a job replaced with a digital better looking, better acting AI replacement. It seem to me the matrix end of human existence is closer then we think.
 
  • Like
Reactions: VitalyT

havok585

TS Addict
1984 is almost here.

The psychopaths that occupy the highest positions on the world stage will use this tech to incriminate anyone they desire.
 
  • Like
Reactions: syrious01

DongBeetle

TS Rookie
If anyone should be scared of this, it's actors soon you all will be out of a job replaced with a digital better looking, better acting AI replacement. It seem to me the matrix end of human existence is closer then we think.
That will never happen because you'll still actors to do the voices.
 
R

Right side bob

Try using the website and look at the people in the background.
 

GamerNerves

TS Rookie
Technology like this can be used to fight crime when trying to recreate possible suspects from only a short voice sample or such. The possibilities for good purposes are numerous. Video games can always use more advanced procedural generation what will benefit from this so called "AI tech". On the other hand, scams and such are going to be rampant.

While masses think that internet should not be controlled in any manner, after a time spent with wildly spread scams and fake computer generated content it is likely that stricter control is desired. Even now some controlling of web anonymity would be positive, so people couldn't shout insults and write false information on the internet without any responsibility.

Scamming people will become more difficult though, since awareness is going to rise at a same phase. People will more likely to check sources and even various tools can be used to scan material that could be generated to know if it is real. As an example antivirus suites can include some tools for this purpose in a near future.

If we take the USA's first manned moonlanding as an example, either if it did really happen or not, mere video footage like @Misagt said above is not enough to proove things anymore. It is possible if fast travelling like flying becomes even more affordable and less of a hassle, people will go more to places themselves and see things with their own eyes. I don't see much popularity for VR in things like museum trips, if even gaming is going to be a breakthrough, since you never cannot be quite sure how real things would look in real life and you simply cannot touch things, which is important.

All in all some interesting possibilities do rise, but careful examination of any information we receive is a matter already and is going get more alerting from now on.
 

Bp968

TS Booster
I think this was written by an AI...

Technology like this can be used to fight crime when trying to recreate possible suspects from only a short voice sample or such. The possibilities for good purposes are numerous. Video games can always use more advanced procedural generation what will benefit from this so called "AI tech". On the other hand, scams and such are going to be rampant.

While masses think that internet should not be controlled in any manner, after a time spent with wildly spread scams and fake computer generated content it is likely that stricter control is desired. Even now some controlling of web anonymity would be positive, so people couldn't shout insults and write false information on the internet without any responsibility.

Scamming people will become more difficult though, since awareness is going to rise at a same phase. People will more likely to check sources and even various tools can be used to scan material that could be generated to know if it is real. As an example antivirus suites can include some tools for this purpose in a near future.

If we take the USA's first manned moonlanding as an example, either if it did really happen or not, mere video footage like @Misagt said above is not enough to proove things anymore. It is possible if fast travelling like flying becomes even more affordable and less of a hassle, people will go more to places themselves and see things with their own eyes. I don't see much popularity for VR in things like museum trips, if even gaming is going to be a breakthrough, since you never cannot be quite sure how real things would look in real life and you simply cannot touch things, which is important.

All in all some interesting possibilities do rise, but careful examination of any information we receive is a matter already and is going get more alerting from now on.