5 ms·

It's totally an illusion. Up-scaling is a lie. Colorization is a lie. It's literally missing information that can never be verified. It's like if I used machin

It's totally an illusion. Up-scaling is a lie. Colorization is a lie. It's literally missing information that can never be verified. It's like if I used machine learning to double the length of the federalist papers. It would just be absurd to treat that like it was a real piece of history.

burnte 6y ago
So what?
c0nsumer 6y ago
There's a bit of a difference between interpolating to fill in spaces between pixels with hints from what's next to them and adding whole new words and sentences.
- jacobr1 6y ago
  Right. A better example might be removing every 5th word from a famous work, say one of shakespeare's sonnets. Then run GPT3 on it to interpolate the missing words. I bet a reasonable fraction of would match the correct words. Some would be acceptable substitutions, maybe you lose some fidelity or wit from the original, but it doesn't impact the overall enjoyment or understanding of the work. Then will be some minority of cases where we get something weird or wrong. Maybe a word that matches the general meaning, but is modern slang for example. Or maybe it guessed something that is plausible in context, but is opposite in meaning that changes something substantial in the work. I would expect that last scenario to be very rare as a fraction of total word substitutions, but also quite likely over the course of a whole piece of work. On the other hand, if we were able to render have a new undiscovered play available constructed from some sort of damaged fragment, rather than my hypothetical test ... having something we could read or see performed rather than just analyze academically seems worth the chance of getting some (even important) nuance in the work wrong.
  - jacobwilliamroy 6y ago
    That's actually a worse analogy. You're obscuring what's actually going on here. I was being nice when I said it was like "doubling the length of the federalist papers." In a proper analogy, our fake document would be 24 times the length of the federalist papers. 96% of our document would be completely fabricated. FAKE. At that point you're learning more about the computer than whatever document you're viewing and it's the same for video. Scaling up from 480p to 4k requires the computer to fill in 96% of the information. That's really only good for entertainment value, not serious academic research into a historical period. Computers aren't magic time-traveling machines. If you want 4k reproductions of old films, just scan the actual film with better equipment. If you want colorization, that requires actual investigation and to figure out what colors were in use, and how to map them onto the video data. If no such investigation is possible then we're SHIT OUT OF LUCK. It happens. Move on. Stop revising history for entertainment.
vlunkr 6y ago
> It's like if I used machine learning to double the length of the federalist papers. I don't think that's an appropriate analogy at all. When colorizing or up-scaling, you're making informed guesses to fill in gaps, and even if it's wrong occasionally, it doesn't matter.
- jacobwilliamroy 6y ago
  The process for forming those guesses is the same whether you're generating text or generating pixels. Data is data. Another risk this poses is that video is not as simple as many assume. A naive approach of simply desaturating or downscaling an EiB of video data is not enough to achieve even a 50% level of historical accuracy because that doesn't address the problem of normalizing those processes across different camera models, lenses, film processes, digital scanning processes, codecs etc.
mthoms 6y ago
Translation of historic texts and glyphs is also "a lie" since information (meaning) is lost or distorted during the translation process. Sometimes that distortion is intentional, sometimes not. The point being, as long as we're aware of the source of the (image or text) translation we can decide for ourselves whether that source is reputable enough to trust.
- jacobwilliamroy 6y ago
  We're not that evolved yet. This technology is outpacing our own evolution as is evidenced by all the people in the comments who claim they actually learned more about history by watching doctored videos.
mcv 6y ago
It's more like scanning and OCRing the federalist papers. Or maybe translating old documents to modern English. Is it bad to translate Plato from Greek to modern English? That involves a whole lot more invention than this process.
- jacobwilliamroy 6y ago
  It's less like OCR'ing the federalist papers and more like increasing the length by 2400%. That's what blowing up a video from 480p to 4k entails: fabricating 96% of the pixels based on the 4% which are actually available in the original piece. It really is as useless as I make it sound.

burnte 6y ago

So what?

c0nsumer 6y ago

There's a bit of a difference between interpolating to fill in spaces between pixels with hints from what's next to them and adding whole new words and sentences.

jacobr1 6y ago
Right. A better example might be removing every 5th word from a famous work, say one of shakespeare's sonnets. Then run GPT3 on it to interpolate the missing words. I bet a reasonable fraction of would match the correct words. Some would be acceptable substitutions, maybe you lose some fidelity or wit from the original, but it doesn't impact the overall enjoyment or understanding of the work. Then will be some minority of cases where we get something weird or wrong. Maybe a word that matches the general meaning, but is modern slang for example. Or maybe it guessed something that is plausible in context, but is opposite in meaning that changes something substantial in the work. I would expect that last scenario to be very rare as a fraction of total word substitutions, but also quite likely over the course of a whole piece of work. On the other hand, if we were able to render have a new undiscovered play available constructed from some sort of damaged fragment, rather than my hypothetical test ... having something we could read or see performed rather than just analyze academically seems worth the chance of getting some (even important) nuance in the work wrong.
- jacobwilliamroy 6y ago
  That's actually a worse analogy. You're obscuring what's actually going on here. I was being nice when I said it was like "doubling the length of the federalist papers." In a proper analogy, our fake document would be 24 times the length of the federalist papers. 96% of our document would be completely fabricated. FAKE. At that point you're learning more about the computer than whatever document you're viewing and it's the same for video. Scaling up from 480p to 4k requires the computer to fill in 96% of the information. That's really only good for entertainment value, not serious academic research into a historical period. Computers aren't magic time-traveling machines. If you want 4k reproductions of old films, just scan the actual film with better equipment. If you want colorization, that requires actual investigation and to figure out what colors were in use, and how to map them onto the video data. If no such investigation is possible then we're SHIT OUT OF LUCK. It happens. Move on. Stop revising history for entertainment.

vlunkr 6y ago

> It's like if I used machine learning to double the length of the federalist papers. I don't think that's an appropriate analogy at all. When colorizing or up-scaling, you're making informed guesses to fill in gaps, and even if it's wrong occasionally, it doesn't matter.

jacobwilliamroy 6y ago
The process for forming those guesses is the same whether you're generating text or generating pixels. Data is data. Another risk this poses is that video is not as simple as many assume. A naive approach of simply desaturating or downscaling an EiB of video data is not enough to achieve even a 50% level of historical accuracy because that doesn't address the problem of normalizing those processes across different camera models, lenses, film processes, digital scanning processes, codecs etc.

mthoms 6y ago

Translation of historic texts and glyphs is also "a lie" since information (meaning) is lost or distorted during the translation process. Sometimes that distortion is intentional, sometimes not. The point being, as long as we're aware of the source of the (image or text) translation we can decide for ourselves whether that source is reputable enough to trust.

jacobwilliamroy 6y ago
We're not that evolved yet. This technology is outpacing our own evolution as is evidenced by all the people in the comments who claim they actually learned more about history by watching doctored videos.

mcv 6y ago

It's more like scanning and OCRing the federalist papers. Or maybe translating old documents to modern English. Is it bad to translate Plato from Greek to modern English? That involves a whole lot more invention than this process.

jacobwilliamroy 6y ago
It's less like OCR'ing the federalist papers and more like increasing the length by 2400%. That's what blowing up a video from 480p to 4k entails: fabricating 96% of the pixels based on the 4% which are actually available in the original piece. It really is as useless as I make it sound.