8 ms·

I assure you that's not enough. These are high quality videos. Once they get uploaded to social media, compression mostly makes imperfections go away. And it's

by lopis 2y ago

I assure you that's not enough. These are high quality videos. Once they get uploaded to social media, compression mostly makes imperfections go away. And it's been shown that when people are not expecting AI content, they are much less likely to realize they are looking at AI. I would 100% believe most of these videos were real if caught off guard.

ddtaylor 2y ago
I regularly catch my kids watching AI generated content and they don't know it.
- throwup238 2y ago
  A surprising amount of it is really popular too. I recently figured out that the Movie Recaps channel was all AI when the generated voice slipped and mispronounced a word in a really unnatural way. They post videos almost daily and they get millions of views. Must be making bank.
  - inkcapmushroom 2y ago
    I was watching UFC recaps on Youtube and the algorithm got me onto AI generated MMA content, I watched for a while before realizing it. They were using old videos which were "enhanced" using AI and had an AI narrator. I only realized it when the fight footage got so old, and the AI had to do so much work to touch it up, that artifacts started appearing in the video. Once I realized it I rewatched the earlier clips in the video and could see the artifacts there too, but not until I was looking for them.
    - araes 2y ago
      There's already rabbit holes of fake MMA fighting you can fall into online? Even if you're a "fan" and relatively aware of what to look for ... still difficult to spot? Horribly, had the same sensation while watching UFC at a bar. "Haven't I seen this match where they fall on the ground and hug for hours before?" Mostly empty background audience with limited reactions. Somebody took AI video editing, and in a year or two, we're already at entire MMA rabbit holes of fake videos. Commenting mostly as a personal evidence reference of how crazy the World Wide Web has gotten from anecdotal sources.
  - Twisell 2y ago
    Most probably they employ overseas, underpaid workers with non-standard English accents and so they include text-to-speach in the production process to smoothen the end result. I won't argue wether text to speech qualifies as an AI but I agree they must be making bank.
    - asveikau 2y ago
      I think it's unusual to assume they are based in the US and employ/underpay foreigners. A lot of people making the content are just from somewhere else.
    - derefr 2y ago
      > Most probably they employ overseas, underpaid workers with non-standard English accents and so they include text-to-speach in the production process to smoothen the end result. Might also be an AI voice-changer (i.e. speech2speech) model. These models are most well-known for being used to create "if [famous singer] performed [famous song not by them]" covers — you sing the song yourself, then run your recording through the model to convert the recording into an equivalent performance in the singer's voice; and then you composite that onto a vocal-less version of the track. But you can just as well use such a model to have overseas workers read a script, and then convert that recording into an "equivalent performance" in a fluent English speaker's voice. Such models just slip up when they hit input phonemes they can't quite understand the meaning of. (If you were setting this up for your own personal use, you could fine-tune the speech2speech model like a translation model, so it understands how your specific accent should map to the target. [I.e., take a bunch of known sample outputs, and create paired inputs by recording your own performances of them.] This wouldn't be tenable for a big low-cost operation, of course, as the recordings would come from temp workers all over the world with high churn.)
      - spywaregorilla 2y ago
        Can you identify any of these models?
    - bee_rider 2y ago
      I wonder if they are making bank. Seems like a race to the bottom, there’s no barrier to entry, right?
  - mystifyingpoi 2y ago
    But it uses AI only for audio, right? Script for the vid seems to be written by human, given the unusual humor type of this channel. I started watching this channel some time ago.
    - throwup238 2y ago
      It's hard to tell whether they use AI for script generation. After having seen enough of those recaps, the humor seems to be rather mechanical and basic humor is relatively easy to get from an LLM if prompted correctly. The video titles also seem as if they were generated. That said, this channel has been producing videos well before ChatGPT3.5/4 so at the very least they probably started with human written scripts.
      - mewpmewp2 2y ago
        I thought it was just text to speech when I happen to saw some of those videos. And it seems to have been consistently similar since before ChatGPT etc. Why do you think titles are AI generated? I feel like it might actually be quite complex for AI to pull up the perfect clips and edit them with the script, including timing and everything. Maybe it could be made automatic, but nonetheless it would be a complex process and I don't think possible few years ago. I know Gemini and possibly some others can analyze video if fed to them, but I'm still skeptical that this channel in particular would have done it, when they have always had this frequency of uploads and similar tone. Also I think there's far better TTS now with ElevenLabs and others so it could be made much more human like.
  - the_af 2y ago
    A group I follow about hobby/miniatures (as in wargaming miniatures and dioramas) recently shared an "awesome" image of a diorama from another "hobby" group. The image had all the telltale signs of being AI generated (too much detail, the lights & shadows were the wrong scale, the focus of the lens was odd for the kind of photo, etc). I checked that other group, and sure enough, they claim to be about sharing "miniature dioramas" but all they share is AI-generated crap. And in the original group, which I'm a member of and is full of people who actually create dioramas -- let's say they are "subject matter experts" -- nobody suspected anything! To them, who are unfamiliar with AI art, the photo was of a real hand-made diorama.
- dham 2y ago
  It's kind of an interesting phenomenon. I read something on this. Basically being born between ~1980 and ~1990 is a huge advantage in tech.
  - ben_w 2y ago
    The only generation that ever knew how to set the clock on a VCR: our parents needed our help; our kids won't have even seen a VCR outside of a museum, much less used one.
    - wsintra2022 2y ago
      Very interesting point. Wonder about the generation before and what skills they had to share with their parents who were most likely traumatised from a world war or two. I remember setting the vcr clock and tuning the new tv with the remote. I’m sure the adults could of figured it out but they probably got more from seeing their ‘smart’ kids figuring it out in time for the cartoons!
    - driverdan 2y ago
      The parents of those of us who grew up in the 80's and 90's invented the VCR, they could use it just fine.
  - WhyOhWhyQ 2y ago
    The Zoomers have the advantage that the bar is pretty low these days.
mikae1 2y ago
> it's been shown that when people are not expecting AI content, they are much less likely to realize they are looking at AI. At this point, looking at a big tech SoMe feed I would expect that everything is, or at least could be, gen AI content.
jetrink 2y ago
A friend who lives in North Carolina sent me a video of the raging floodwaters in his state- at least that's what the superimposed text claimed it was. When I looked closer, it was clearly an Indian city filled with Indian people and Indian cars. He hadn't noticed anything except the flood water. It reminded me of that famous selective attention test video[1]. I won't ruin it for those who haven't seen it, but it's amazing what details we can miss when we aren't looking for them. I suspect this is made even worse when we're casually viewing videos in a disjointed way as on social media and we're not even giving one part of the video our full attention. 1. https://www.youtube.com/watch?v=vJG698U2Mvo https://www.youtube.com/watch?v=vJG698U2Mvo
- itslennysfault 2y ago
  hmmm... Maybe it's because I knew it was testing me, but I noticed it right away and counted the right count. I could see it being pretty shocking if I hadn't, but I honestly can't imagine how I'd miss that.
  - firebaze 2y ago
    > hmmm... Maybe it's because I knew it was testing me, but I noticed it right away and counted the right count. > I could see it being pretty shocking if I hadn't, but I honestly can't imagine how I'd miss that. The point of the video wasn't to count correctly, but to see the gorilla
    - yunwal 2y ago
      cool, he noticed it right away
      - andrewinardeer 2y ago
        I believe them. Why would people lie on the internet?
    - bee_rider 2y ago
      99% the person was playing along for the rest of us, so we get a chance to enjoy the video as intended.
valval 2y ago
The way I see it, it won’t take long before human eyes won’t be able to distinguish AI generated content from original. The only regret I have about that is losing video as a form of evidence. CCTV footage and the like are a valuable tool for solving crimes. That’s going to be out the window soon.
- akoboldfrying 2y ago
  Trust can be preserved by adding PKI at the hardware level. What you said about CCTV is true; once the market realises and demand appears, camera manufacturers will start making camera modules that, e.g., sign each frame with the manufacturer's private key, enabling Joe Public to verify that that frame came from a camera made by that manufacturer. Reputational risk makes the manufacturer store the private key in the device in a secure, tamper-proof way (like TPMs do now), which (mostly) prevents those private keys from leaking. Does this create difficulties if you want to modify the raw video data in any way? Yes it does, even if you just want to save it in a different lossy compression level or format. But these problems aren't insurmountable. Essentially, provenance info can be added for each modification, signed by the entity that made the change, and the end viewer can then decide if they trust the full certificate chain (just as they do now with HTTPS).
  - gitaarik 2y ago
    Oh wow, that's a great idea. Isn't this already happening maybe? Recently someone said here that it's noticable that videos from CCTV cameras are often filmed with a phone or camera on a screen instead of using the original video, and people were speculating that it might be hard or impossible to get access to the original recording because of bureaucracy or something, but that recording a playback on a screen with a phone or camera or something is then often allowed. Maybe they also do this partly so that the original can't be easily messed with by other people. But yeah if you can verify that a certain video was filmed at a certain time by a certain camera, that is great. Of course the companies providing these cameras should be trustworthy, and that the camera's are actually really sending what they actually record, and that the company itself doesn't mess with the original recordings.
mikehollinger 2y ago
> compression mostly makes imperfections go away The ultimate compression is to reduce the video clip to a latent space vector representation to be rendered on device. :) Just give us a few more revs of Moore’s law for that to be reasonable. edit: found a patent… https://patents.google.com/patent/US11388416B2/en https://patents.google.com/patent/US11388416B2/en

by lopis 2y ago

ddtaylor 2y ago

I regularly catch my kids watching AI generated content and they don't know it.

throwup238 2y ago
A surprising amount of it is really popular too. I recently figured out that the Movie Recaps channel was all AI when the generated voice slipped and mispronounced a word in a really unnatural way. They post videos almost daily and they get millions of views. Must be making bank.
- inkcapmushroom 2y ago
  I was watching UFC recaps on Youtube and the algorithm got me onto AI generated MMA content, I watched for a while before realizing it. They were using old videos which were "enhanced" using AI and had an AI narrator. I only realized it when the fight footage got so old, and the AI had to do so much work to touch it up, that artifacts started appearing in the video. Once I realized it I rewatched the earlier clips in the video and could see the artifacts there too, but not until I was looking for them.
  - araes 2y ago
    There's already rabbit holes of fake MMA fighting you can fall into online? Even if you're a "fan" and relatively aware of what to look for ... still difficult to spot? Horribly, had the same sensation while watching UFC at a bar. "Haven't I seen this match where they fall on the ground and hug for hours before?" Mostly empty background audience with limited reactions. Somebody took AI video editing, and in a year or two, we're already at entire MMA rabbit holes of fake videos. Commenting mostly as a personal evidence reference of how crazy the World Wide Web has gotten from anecdotal sources.
- Twisell 2y ago
  Most probably they employ overseas, underpaid workers with non-standard English accents and so they include text-to-speach in the production process to smoothen the end result. I won't argue wether text to speech qualifies as an AI but I agree they must be making bank.
  - asveikau 2y ago
    I think it's unusual to assume they are based in the US and employ/underpay foreigners. A lot of people making the content are just from somewhere else.
  - derefr 2y ago
    > Most probably they employ overseas, underpaid workers with non-standard English accents and so they include text-to-speach in the production process to smoothen the end result. Might also be an AI voice-changer (i.e. speech2speech) model. These models are most well-known for being used to create "if [famous singer] performed [famous song not by them]" covers — you sing the song yourself, then run your recording through the model to convert the recording into an equivalent performance in the singer's voice; and then you composite that onto a vocal-less version of the track. But you can just as well use such a model to have overseas workers read a script, and then convert that recording into an "equivalent performance" in a fluent English speaker's voice. Such models just slip up when they hit input phonemes they can't quite understand the meaning of. (If you were setting this up for your own personal use, you could fine-tune the speech2speech model like a translation model, so it understands how your specific accent should map to the target. [I.e., take a bunch of known sample outputs, and create paired inputs by recording your own performances of them.] This wouldn't be tenable for a big low-cost operation, of course, as the recordings would come from temp workers all over the world with high churn.)
    - spywaregorilla 2y ago
      Can you identify any of these models?
  - bee_rider 2y ago
    I wonder if they are making bank. Seems like a race to the bottom, there’s no barrier to entry, right?
- mystifyingpoi 2y ago
  But it uses AI only for audio, right? Script for the vid seems to be written by human, given the unusual humor type of this channel. I started watching this channel some time ago.
  - throwup238 2y ago
    It's hard to tell whether they use AI for script generation. After having seen enough of those recaps, the humor seems to be rather mechanical and basic humor is relatively easy to get from an LLM if prompted correctly. The video titles also seem as if they were generated. That said, this channel has been producing videos well before ChatGPT3.5/4 so at the very least they probably started with human written scripts.
    - mewpmewp2 2y ago
      I thought it was just text to speech when I happen to saw some of those videos. And it seems to have been consistently similar since before ChatGPT etc. Why do you think titles are AI generated? I feel like it might actually be quite complex for AI to pull up the perfect clips and edit them with the script, including timing and everything. Maybe it could be made automatic, but nonetheless it would be a complex process and I don't think possible few years ago. I know Gemini and possibly some others can analyze video if fed to them, but I'm still skeptical that this channel in particular would have done it, when they have always had this frequency of uploads and similar tone. Also I think there's far better TTS now with ElevenLabs and others so it could be made much more human like.
- the_af 2y ago
  A group I follow about hobby/miniatures (as in wargaming miniatures and dioramas) recently shared an "awesome" image of a diorama from another "hobby" group. The image had all the telltale signs of being AI generated (too much detail, the lights & shadows were the wrong scale, the focus of the lens was odd for the kind of photo, etc). I checked that other group, and sure enough, they claim to be about sharing "miniature dioramas" but all they share is AI-generated crap. And in the original group, which I'm a member of and is full of people who actually create dioramas -- let's say they are "subject matter experts" -- nobody suspected anything! To them, who are unfamiliar with AI art, the photo was of a real hand-made diorama.
dham 2y ago
It's kind of an interesting phenomenon. I read something on this. Basically being born between ~1980 and ~1990 is a huge advantage in tech.
- ben_w 2y ago
  The only generation that ever knew how to set the clock on a VCR: our parents needed our help; our kids won't have even seen a VCR outside of a museum, much less used one.
  - wsintra2022 2y ago
    Very interesting point. Wonder about the generation before and what skills they had to share with their parents who were most likely traumatised from a world war or two. I remember setting the vcr clock and tuning the new tv with the remote. I’m sure the adults could of figured it out but they probably got more from seeing their ‘smart’ kids figuring it out in time for the cartoons!
  - driverdan 2y ago
    The parents of those of us who grew up in the 80's and 90's invented the VCR, they could use it just fine.
- WhyOhWhyQ 2y ago
  The Zoomers have the advantage that the bar is pretty low these days.

mikae1 2y ago

> it's been shown that when people are not expecting AI content, they are much less likely to realize they are looking at AI. At this point, looking at a big tech SoMe feed I would expect that everything is, or at least could be, gen AI content.

jetrink 2y ago

A friend who lives in North Carolina sent me a video of the raging floodwaters in his state- at least that's what the superimposed text claimed it was. When I looked closer, it was clearly an Indian city filled with Indian people and Indian cars. He hadn't noticed anything except the flood water. It reminded me of that famous selective attention test video[1]. I won't ruin it for those who haven't seen it, but it's amazing what details we can miss when we aren't looking for them. I suspect this is made even worse when we're casually viewing videos in a disjointed way as on social media and we're not even giving one part of the video our full attention. 1. https://www.youtube.com/watch?v=vJG698U2Mvo https://www.youtube.com/watch?v=vJG698U2Mvo

itslennysfault 2y ago
hmmm... Maybe it's because I knew it was testing me, but I noticed it right away and counted the right count. I could see it being pretty shocking if I hadn't, but I honestly can't imagine how I'd miss that.
- firebaze 2y ago
  > hmmm... Maybe it's because I knew it was testing me, but I noticed it right away and counted the right count. > I could see it being pretty shocking if I hadn't, but I honestly can't imagine how I'd miss that. The point of the video wasn't to count correctly, but to see the gorilla
  - yunwal 2y ago
    cool, he noticed it right away
    - andrewinardeer 2y ago
      I believe them. Why would people lie on the internet?
  - bee_rider 2y ago
    99% the person was playing along for the rest of us, so we get a chance to enjoy the video as intended.

valval 2y ago

The way I see it, it won’t take long before human eyes won’t be able to distinguish AI generated content from original. The only regret I have about that is losing video as a form of evidence. CCTV footage and the like are a valuable tool for solving crimes. That’s going to be out the window soon.

akoboldfrying 2y ago
Trust can be preserved by adding PKI at the hardware level. What you said about CCTV is true; once the market realises and demand appears, camera manufacturers will start making camera modules that, e.g., sign each frame with the manufacturer's private key, enabling Joe Public to verify that that frame came from a camera made by that manufacturer. Reputational risk makes the manufacturer store the private key in the device in a secure, tamper-proof way (like TPMs do now), which (mostly) prevents those private keys from leaking. Does this create difficulties if you want to modify the raw video data in any way? Yes it does, even if you just want to save it in a different lossy compression level or format. But these problems aren't insurmountable. Essentially, provenance info can be added for each modification, signed by the entity that made the change, and the end viewer can then decide if they trust the full certificate chain (just as they do now with HTTPS).
- gitaarik 2y ago
  Oh wow, that's a great idea. Isn't this already happening maybe? Recently someone said here that it's noticable that videos from CCTV cameras are often filmed with a phone or camera on a screen instead of using the original video, and people were speculating that it might be hard or impossible to get access to the original recording because of bureaucracy or something, but that recording a playback on a screen with a phone or camera or something is then often allowed. Maybe they also do this partly so that the original can't be easily messed with by other people. But yeah if you can verify that a certain video was filmed at a certain time by a certain camera, that is great. Of course the companies providing these cameras should be trustworthy, and that the camera's are actually really sending what they actually record, and that the company itself doesn't mess with the original recordings.

mikehollinger 2y ago

> compression mostly makes imperfections go away The ultimate compression is to reduce the video clip to a latent space vector representation to be rendered on device. :) Just give us a few more revs of Moore’s law for that to be reasonable. edit: found a patent… https://patents.google.com/patent/US11388416B2/en https://patents.google.com/patent/US11388416B2/en