An AI edited photo might not necessarily be less representative of whatever is in the photo. Imagine an image taken in a very dark room, then an AI enhancement makes it look like the lights are on. You can actually get a much better idea of what’s in the room, but a less good idea of what the lighting was like. So it comes down to opinion, which one is more representative of reality? Because no photo since the beginning of time has been completely representative of what humans actually see with their eyes. It’s always been a trade-off of: what do we change to give humans the image they want with the technology we have.
Do you think night vision produces a ‘fake’ image? Maybe you do, but my point is, that’s your opinion. You might think that accurate representation of the light level is more important than accurate representation of the objects in front of the lens. But someone else might not. Same way a colorized photo can give a more accurate representation of reality with false information.
I mean, you’re debating the meaning of “accurate representation”. We may as well debate the meaning of perception, too, but I don’t think it changes the point of my original argument.
I think it does, because photos have always been an inaccurate representation of what a person sees. You zoom in on my face in a picture and you see a bunch of pixels. That’s not what my face looks like, I’m not made of tiny boxes. If I AI upscale it, it looks a lot closer. My argument here is simply: the statement that an AI dependent image is inherently less representative of reality, is not necessarily true.
An AI edited photo might not necessarily be less representative of whatever is in the photo. Imagine an image taken in a very dark room, then an AI enhancement makes it look like the lights are on. You can actually get a much better idea of what’s in the room, but a less good idea of what the lighting was like. So it comes down to opinion, which one is more representative of reality? Because no photo since the beginning of time has been completely representative of what humans actually see with their eyes. It’s always been a trade-off of: what do we change to give humans the image they want with the technology we have.
…but the lights weren’t on.
Do you think night vision produces a ‘fake’ image? Maybe you do, but my point is, that’s your opinion. You might think that accurate representation of the light level is more important than accurate representation of the objects in front of the lens. But someone else might not. Same way a colorized photo can give a more accurate representation of reality with false information.
I mean, you’re debating the meaning of “accurate representation”. We may as well debate the meaning of perception, too, but I don’t think it changes the point of my original argument.
I think it does, because photos have always been an inaccurate representation of what a person sees. You zoom in on my face in a picture and you see a bunch of pixels. That’s not what my face looks like, I’m not made of tiny boxes. If I AI upscale it, it looks a lot closer. My argument here is simply: the statement that an AI dependent image is inherently less representative of reality, is not necessarily true.