SomeMichGuy · M
The first ones were easy to spot.
They're getting better.
I often refer to the visual thing as an "image" if I think it's AI.
The crazy thing is that AI "detectors" can violently disagree about some images (one will say nearly 100% AI-generated, another will say nearly 100% real).
They're getting better.
I often refer to the visual thing as an "image" if I think it's AI.
The crazy thing is that AI "detectors" can violently disagree about some images (one will say nearly 100% AI-generated, another will say nearly 100% real).
CountScrofula · 41-45, M
Frostcloud · F
that may have been true a few years ago, but it's actually everywhere now. a lot of it is shown as "ring camera" footage for example, because it's easier to pass real, since doorbell cameras wont have the best sound or videk quality. but there's high quality ai videos out there too
swirlie · 31-35
The telltale sign of an AI image is it's distinct lack of 3-dimensional imagery perspective. This means that an AI computer-rendering of an image is incapable of simulating 'depth perception' of an object or a scene. This is because AI does not use 3-dimensional imaging.
Even your cellphone camera is incapable of depicting 'depth perception' like the human eye can otherwise do (assuming two eyes are being used) because your cell camera uses 2-dimensional technology as do most standard cameras today.
That said, the only option open to depict depth for an AI-generated image is through the use of pre-computed shadowing techniques to suggest some level of depth or difference exists when compared to a surface area which has no shadows.
A standard camera uses this same technique to show depth, tough that sense of depth is only what you know to be true from within your own experiences. Therefore, when you see a shadow within a camera image, you know from experience that one object is further away from the other in real life, so your mind's eye sees it that way in the 2D photo as well.
Relative to a flat surface area, a standard camera will capture shadows which might suggest that something is deeper than or further away from the reference point of the surface area.
An AI computer-generated image however, uses only 1-dimensional imaging (flat imaging) and will attempt to use natural light and pre-known experiences that are built-in to the computer model which represents 'depth' to depict some semblance of depth perception within an AI image. Any AI-generated image which depicts depth will have had that pre-known feature built into the computer's AI logic.
An example of this would be standing on the edge of the Grand Canyon and looking straight down with your eyes. You would be able to rationalize depth within the canyon but your camera cannot do that, nor can an AI image create the illusion of depth unless that perspective is programed into the AI generator.
Even your cellphone camera is incapable of depicting 'depth perception' like the human eye can otherwise do (assuming two eyes are being used) because your cell camera uses 2-dimensional technology as do most standard cameras today.
That said, the only option open to depict depth for an AI-generated image is through the use of pre-computed shadowing techniques to suggest some level of depth or difference exists when compared to a surface area which has no shadows.
A standard camera uses this same technique to show depth, tough that sense of depth is only what you know to be true from within your own experiences. Therefore, when you see a shadow within a camera image, you know from experience that one object is further away from the other in real life, so your mind's eye sees it that way in the 2D photo as well.
Relative to a flat surface area, a standard camera will capture shadows which might suggest that something is deeper than or further away from the reference point of the surface area.
An AI computer-generated image however, uses only 1-dimensional imaging (flat imaging) and will attempt to use natural light and pre-known experiences that are built-in to the computer model which represents 'depth' to depict some semblance of depth perception within an AI image. Any AI-generated image which depicts depth will have had that pre-known feature built into the computer's AI logic.
An example of this would be standing on the edge of the Grand Canyon and looking straight down with your eyes. You would be able to rationalize depth within the canyon but your camera cannot do that, nor can an AI image create the illusion of depth unless that perspective is programed into the AI generator.
View 28 more replies »
idontcareok · 70-79, M
@swirlie seems, that i won't be able to fool you with ai pics of me, lol
bijouxbroussard · F
Hannahjay14 · F
WoWgirll · 36-40, F
ChipmunkErnie · 70-79, M
ChipmunkErnie · 70-79, M
@TradEmily If you had been, how would you know? Wouldn't that be the definition of "fooled"?
OogieBoogie · F