I hate what this means for the world. It accelerates the circling of the dystopian drain. Makes me feel that i might as well spend retirement funds on getting shitfaced, because that's not a world...
I hate what this means for the world. It accelerates the circling of the dystopian drain. Makes me feel that i might as well spend retirement funds on getting shitfaced, because that's not a world I want to see, let alone live in.
I was making a Cold War Space Race display for my classroom the other day and kept coming across AI-generated images, as well as images that were just mislabeled. It was very difficult to sort...
I was making a Cold War Space Race display for my classroom the other day and kept coming across AI-generated images, as well as images that were just mislabeled. It was very difficult to sort through what was real and I ended up having to rely on the BBC and a few other centralized/trusted sources.
It made me realize we are probably pretty close to the point where history becomes a really huge problem. It's easy to verify the Holocaust right now but in just a few short decades there won't be any living people who have even met a survivor. What happens when AI creates so much bullshit that you can't even tell whether a book was actually published before the invention of AI? What happens when a museum or credible news organization mistakenly adds a fake WW2 photo to their virtual tour and people start to wonder if any of it really happened?
I suspect there will be a solution - probably some combination of digital signatures and harsh laws. But I'm not excited about what happens in the meantime.
If only we lived in a world where these advances would be used to make everyone's lives better; instead of putting millions of people out of work and then hoarding those savings for the...
If only we lived in a world where these advances would be used to make everyone's lives better; instead of putting millions of people out of work and then hoarding those savings for the billionaires (soon to be trillionaires.)
Human productivity has skyrocketed in the last 100 years yet many live in complete poverty and most cannot afford homes, healthcare, or even healthy food. We live in a world of the "have and the have nots" where you either have wealth or you don't. Hell, "middle class" people can't even afford to buy stupid luxuries with their meager wages anymore.
Oof, these are getting disturbingly good. For the most part, AI-generated videos usually jump right out to me, but that’s slipping away fast. I was convinced the basketball one was real until I...
Oof, these are getting disturbingly good. For the most part, AI-generated videos usually jump right out to me, but that’s slipping away fast. I was convinced the basketball one was real until I noticed the messed up text everywhere.
I like to think I’d’ve done better given longer clips.
Very convincing. But I do have to question how hard it is to fake compressed short-form low-quality videos. I still haven't seen any AI-generated videos that seem convincingly like a movie. And of...
Very convincing. But I do have to question how hard it is to fake compressed short-form low-quality videos. I still haven't seen any AI-generated videos that seem convincingly like a movie. And of course the short-form video is a great way to work around the limitations of limited context and persistence; right now, if a video spins 360 degrees and the scene stays consistent throughout, you know 99% confidence that a video is real. The same thing applies to longer form content in general; it's easier to spot inconsistencies if a scene goes on long enough that inconsistencies can actually happen. 10 seconds is simply not enough time.
This is really really really scary considering how many people consume tons of short-form video. Many people do not consume content through a critical lens. Many people get news from short-form videos on TikTok and Instagram. I've already noticed that these people essentially live in a different universe from myself, believing whatever the hottest algo influencers are pushing in a given week. Surely this will make that worse, and easier to manipulate than ever -- no need for companies, governments, etc. to pay off an influencer to create propaganda for you when you can just generate it yourself through a hot influencer-ish AI avatar.
I continue to feel that short-form video is the lowest form of communication content available today, right next to image memes. Not enough length to give citations or make a coherent argument with any nuance. Reliant on eye-grabbing techniques just like YouTube previews. I call these 'eyeworms' because they remind me of music earworms, and yeah, I want the term to sound disgusting. But honestly this stuff is getting creepily similar to TV advertisements -- no substance, just trying to sell you something or brainwash you in 10-30 seconds. And unlike TV ads, there's no regulation whatsoever. So you don't even get a giant screen of tiny text sideeffects read in a super fast voice when someone tries to sell you drugs on TikTok.
Back to RSS feeds and blogs and obsolescence I go.
I'm working on a hardware product at the moment. I mentioned it to someone at a party in San Francisco and their first reaction was I should create videos using Veo 3 to market the device. Why? If...
I'm working on a hardware product at the moment. I mentioned it to someone at a party in San Francisco and their first reaction was I should create videos using Veo 3 to market the device. Why? If I have the device why wouldn't I just record it as I hold it in my hand? We're in for some tasteless misuse from people like him in the near future. That happens any time constraints are significantly lifted from the creative process. Compare video game graphics from the SNES age and the first generation of consoles with 3D and full color range. Most of the 3D games were lost in the new expansive possibilities, while the last generation of hardware that mandated 2D graphics kept options limited and channeled creativity. It took a while for people to learn what artificial constraints were necessary to make new games look good.
I spent about a minute or so on each of the videos and got 7/10. I only felt confident in three of my answers. The short clips make it harder to detect some of the weird things that AI videos are...
I spent about a minute or so on each of the videos and got 7/10. I only felt confident in three of my answers. The short clips make it harder to detect some of the weird things that AI videos are known for, but I don't know how much better I'd do if they were longer. I definitely wouldn't have noticed if I were just scrolling though Instagram and not actively looking for AI.
I also got 7/10. Two of the ones I got wrong I was really undecided on, but still chose wrong in the end. The newscasting one felt the most difficult, due to the translucent desk which really made...
I also got 7/10. Two of the ones I got wrong I was really undecided on, but still chose wrong in the end. The newscasting one felt the most difficult, due to the translucent desk which really made it difficult to trace the light/reflections.
Interestingly, a few jumped out to me as obviously real or obviously AI within just a few frames. Others I had to study intensely. It'll be interesting to study what these "tells" are (lighting, sound, compression artifacts?), and how they adapt over time.
AI trained visual clues based detection will be used to fuel better AI videos immediately. We have now entered into a new age. What can we use now to break out? Constant video recording from each...
AI trained visual clues based detection will be used to fuel better AI videos immediately.
We have now entered into a new age. What can we use now to break out? Constant video recording from each person, so that only videos with plausible continuity (verified by continuity based analysis AI) can be accepted as truth?
No, we go back to the old ways: truth is what everyone in a certain space agrees upon. The period of modernity up to now only has this absurd concept of Truth as ultimate and self-evident thanks...
No, we go back to the old ways: truth is what everyone in a certain space agrees upon.
The period of modernity up to now only has this absurd concept of Truth as ultimate and self-evident thanks to limits on who had access to distributing their ideas. The emperor has his invisible suit because he's the center of power and thus defines truth. Stalin could always alter documents.
However, democratized mass media in the form of the internet has splintered the myth of a single coherent reality already, without convincing AI generated "evidence". We are reasonable creatures, not rational ones with perfect senses, and our internal perceptions have always had to be contextualized, with previous data becoming present assumptions. The reason comes after that stage. We can blame bad actors (Fox, bot farms) spreading misinformation all we like, but the fact is that there has always been a diversity of "realities", even through the peak of centralized information channels. Differences of religion, political ideals, who killed JFK, colour sensitivity, who is attractive, we call these things varying degrees of "acceptably subjective", but they're all fundamental aspects of each subject's conception of the world. This was only possible because of shared beliefs among the gatekeepers of communication.
We need to stop relying on authorities to understand things for us before we can stop relying on them to be honest. We need to stop relying on them to be honest in a world split between fascism (lite?) and Maoism (with market features), because they will never agree, and never have. We collectively avoided the crisis of the death of god solely via inertia and assuming that nobody would do such a thing as tell lies. But now nobody agrees who is the bad guy, so we can't attribute variation in worldviews to bad people.
Edit to add: I'm framing this so black-and-whitely because the only alternatives rely on either a complete revolution into utopia, basically starting tomorrow, or centralized authorities. Even the most trustworthy people and organizations can be co-opted, however. Then we would get to experience what 1984 was really about.
(also, just to be clear, I think Oswald killed Kennedy, and acted alone)
This meshes with my point of view. It feels like humans are no further from the truth than we have been; if anything, we are closer to the truth, if such a thing exists, because we have more...
This meshes with my point of view.
It feels like humans are no further from the truth than we have been; if anything, we are closer to the truth, if such a thing exists, because we have more information.
The information is flowing so much that we can see more of the diversity of viewpoints that have always been there. More humans can see orders of magnitude more viewpoints than ever before, and the information and memes are bouncing around faster.
This could all be of great benefit to us if our nervous systems could adapt to process it properly. But who knows if we'll survive long enough to adapt into a human society that can know that much of itself.
Thanks for bringing up the Second Coming, it's a wonderful balm for these times. One of the few that does not diminish the weight of the pain it soothes. I'm more cynical about even hypothetical...
Thanks for bringing up the Second Coming, it's a wonderful balm for these times. One of the few that does not diminish the weight of the pain it soothes.
I'm more cynical about even hypothetical truth access myself, but it's certainly true that we have more useful models than ever before. Time and time again, people have urged those in power to democratize the responsibility of democracy, to reduce the social focus on production for its own sake, and to deescalate international tensions. The power simply continues to accrue and seems locked into its vector of motion. More and more, I see normal people catch onto this drift, and this is my source of hope. It requires buy-in. It requires us all to subjugate ourselves in the name of how things have to be, as defined by that very subjugation. Nobody can save us, but we can pull out of the dive.
I type this sincerely, directed toward anyone reading it: You are strong and able to survive the vertigo of today and the possible horrors ahead. Your love for others and the efforts it drives will be felt. I love you, and hope you love yourself enough to believe it. That is enough.
Thank you for that last thought, wervenyt, and may it also be with you. I had this thought yesterday that at various points in history, things didn't turn out as poorly as they could have, and...
Thank you for that last thought, wervenyt, and may it also be with you. I had this thought yesterday that at various points in history, things didn't turn out as poorly as they could have, and even in the very worst of it, there are some who survive who go on to love and be loved. We'll be okay because we have each other. Hope does not come from grand changes or radical regime collapse etc, but when we connect with fellow humans one on one, or in very small handfuls, and we're kind to each other and doing small loving hopeful Jung's things together.
Argh, I've been hit! Agreed. And even if we will not be okay despite that love and one another, it's the table stakes. Survival cannot be without either, that is at best death delayed by parody....
also be with you
Argh, I've been hit!
Agreed. And even if we will not be okay despite that love and one another, it's the table stakes. Survival cannot be without either, that is at best death delayed by parody.
(Did you mean to write "things" where it says "Jung's"? If you meant Jung's, that might change our agreement a bit.)
Public key infrastructure that goes down to the firmware in cameras. At least then your average scammer can’t fake a video. Big tech companies or state actors can. But it’s an improvement. Both...
Public key infrastructure that goes down to the firmware in cameras. At least then your average scammer can’t fake a video. Big tech companies or state actors can. But it’s an improvement.
Both Android and iOS devices have most or all of the tech necessary to create attestable images.
It’s important to have as a tool for all sorts of situations. Any time money is involved and you want to verify a photo, passports, in courts, in news. Most photos don’t need to be verifiable. But...
It’s important to have as a tool for all sorts of situations. Any time money is involved and you want to verify a photo, passports, in courts, in news. Most photos don’t need to be verifiable. But some must be.
google have synthid as a edit-resistant watermark on their AI content, and currently have given journalists access to a detector for it, but when competing models get better they're unlikely to...
google have synthid as a edit-resistant watermark on their AI content, and currently have given journalists access to a detector for it, but when competing models get better they're unlikely to have similar watermarks unless this gets legislated on
I'd rather make regulation to disclose when a video is generated mostly with AI. Best way to remove a gift is to make them honest. If people still don't care from there... Well, not much surprises...
I'd rather make regulation to disclose when a video is generated mostly with AI. Best way to remove a gift is to make them honest.
If people still don't care from there... Well, not much surprises me in 2025.
In its current iteration Veo mostly spits out tinny sounding and low quality human voices. It's a pretty obvious tell IMO, though I'm not sure how it translates to lower quality output (e.g. a...
In its current iteration Veo mostly spits out tinny sounding and low quality human voices. It's a pretty obvious tell IMO, though I'm not sure how it translates to lower quality output (e.g. a phone speaker) or how long it will remain so obvious. Ignoring visual inconsistencies, only the makeup tutorial wasn't immediately obvious from the audio to me, because certain women's voices recorded on a low quality microphone (phone) could come out sounding a bit like that.
FWIW I got 10/10 but am in no way discounting that short form AI generated video can be pretty convincing to a large audience these days, especially in the context these would normally be viewed in (uncritically, on a phone, single loop)
I got a 10/10, but I basically did it through audio only. AI sounds like it's being run through a heavy noise reduction filter, and everything has the same audio quality, regardless of how...
I got a 10/10, but I basically did it through audio only. AI sounds like it's being run through a heavy noise reduction filter, and everything has the same audio quality, regardless of how "professional" or "non-professional" the setting is. The videos that were real had reflections, background noise, and different levels of compression that the AI didn't have. I don't think I would have gotten 10/10 on mute, though.
I don't think our scores matter all that much because the point is already made: AI videos are good enough to fool lots of people. This particular test was done in such a way as to try and fool...
I don't think our scores matter all that much because the point is already made: AI videos are good enough to fool lots of people. This particular test was done in such a way as to try and fool you, but it doesn't really matter. Hardly anyone will get 10/10, so it's already a scary future we are in.
I've been saying for a couple of years now that we need transparency regulations. AI absolutely must be labeled clearly as such, and it's scandalous that governments are dragging their heels on this. I know it won't stop people publishing fake videos, but it will be a start and at least stop major brands from doing it.
9/10 Spoilers And honestly, had I looked more critically at at #7. I would have definitely gotten 10/10. The audio in #7 was particularly good, which was the most glaring issue in the other ones....
9/10
Spoilers
And honestly, had I looked more critically at at #7. I would have definitely gotten 10/10. The audio in #7 was particularly good, which was the most glaring issue in the other ones.
Im surprised most people got the crypto bro wrong, which emphasizes how people's reasoning is biased. Only 35% guessed it correctly although the video and audio was perfect!
The false positives are the most interesting part of this imo. My theory is if someone doesn't know what to look for they fall back to what they think the apparent difficulty of faking it using AI...
The false positives are the most interesting part of this imo. My theory is if someone doesn't know what to look for they fall back to what they think the apparent difficulty of faking it using AI would be. A single subject in a common scenario with nothing going on in the background? Dead simple - must be AI.
Interesting, I had a hard time with the audio in a handful of them. Two of my mistakes were falsely accusing real videos, including the one you mentioned.
Interesting, I had a hard time with the audio in a handful of them. Two of my mistakes were falsely accusing real videos, including the one you mentioned.
The videos didn’t work for me from archive, but my library has a 3 day all access pass to New York Times digital. If you keep the page open and log in, you can refresh the original page and it...
The videos didn’t work for me from archive, but my library has a 3 day all access pass to New York Times digital. If you keep the page open and log in, you can refresh the original page and it should work without going through the surveys.
I hate what this means for the world. It accelerates the circling of the dystopian drain. Makes me feel that i might as well spend retirement funds on getting shitfaced, because that's not a world I want to see, let alone live in.
I was making a Cold War Space Race display for my classroom the other day and kept coming across AI-generated images, as well as images that were just mislabeled. It was very difficult to sort through what was real and I ended up having to rely on the BBC and a few other centralized/trusted sources.
It made me realize we are probably pretty close to the point where history becomes a really huge problem. It's easy to verify the Holocaust right now but in just a few short decades there won't be any living people who have even met a survivor. What happens when AI creates so much bullshit that you can't even tell whether a book was actually published before the invention of AI? What happens when a museum or credible news organization mistakenly adds a fake WW2 photo to their virtual tour and people start to wonder if any of it really happened?
I suspect there will be a solution - probably some combination of digital signatures and harsh laws. But I'm not excited about what happens in the meantime.
Every day seems to be a new headline about AI becoming more unstoppable, faster. I'm so very tired.
If only we lived in a world where these advances would be used to make everyone's lives better; instead of putting millions of people out of work and then hoarding those savings for the billionaires (soon to be trillionaires.)
Human productivity has skyrocketed in the last 100 years yet many live in complete poverty and most cannot afford homes, healthcare, or even healthy food. We live in a world of the "have and the have nots" where you either have wealth or you don't. Hell, "middle class" people can't even afford to buy stupid luxuries with their meager wages anymore.
Oof, these are getting disturbingly good. For the most part, AI-generated videos usually jump right out to me, but that’s slipping away fast. I was convinced the basketball one was real until I noticed the messed up text everywhere.
I like to think I’d’ve done better given longer clips.
Very convincing. But I do have to question how hard it is to fake compressed short-form low-quality videos. I still haven't seen any AI-generated videos that seem convincingly like a movie. And of course the short-form video is a great way to work around the limitations of limited context and persistence; right now, if a video spins 360 degrees and the scene stays consistent throughout, you know 99% confidence that a video is real. The same thing applies to longer form content in general; it's easier to spot inconsistencies if a scene goes on long enough that inconsistencies can actually happen. 10 seconds is simply not enough time.
This is really really really scary considering how many people consume tons of short-form video. Many people do not consume content through a critical lens. Many people get news from short-form videos on TikTok and Instagram. I've already noticed that these people essentially live in a different universe from myself, believing whatever the hottest algo influencers are pushing in a given week. Surely this will make that worse, and easier to manipulate than ever -- no need for companies, governments, etc. to pay off an influencer to create propaganda for you when you can just generate it yourself through a hot influencer-ish AI avatar.
I continue to feel that short-form video is the lowest form of communication content available today, right next to image memes. Not enough length to give citations or make a coherent argument with any nuance. Reliant on eye-grabbing techniques just like YouTube previews. I call these 'eyeworms' because they remind me of music earworms, and yeah, I want the term to sound disgusting. But honestly this stuff is getting creepily similar to TV advertisements -- no substance, just trying to sell you something or brainwash you in 10-30 seconds. And unlike TV ads, there's no regulation whatsoever. So you don't even get a giant screen of tiny text sideeffects read in a super fast voice when someone tries to sell you drugs on TikTok.
Back to RSS feeds and blogs and obsolescence I go.
I'm working on a hardware product at the moment. I mentioned it to someone at a party in San Francisco and their first reaction was I should create videos using Veo 3 to market the device. Why? If I have the device why wouldn't I just record it as I hold it in my hand? We're in for some tasteless misuse from people like him in the near future. That happens any time constraints are significantly lifted from the creative process. Compare video game graphics from the SNES age and the first generation of consoles with 3D and full color range. Most of the 3D games were lost in the new expansive possibilities, while the last generation of hardware that mandated 2D graphics kept options limited and channeled creativity. It took a while for people to learn what artificial constraints were necessary to make new games look good.
I spent about a minute or so on each of the videos and got 7/10. I only felt confident in three of my answers. The short clips make it harder to detect some of the weird things that AI videos are known for, but I don't know how much better I'd do if they were longer. I definitely wouldn't have noticed if I were just scrolling though Instagram and not actively looking for AI.
I also got 7/10. Two of the ones I got wrong I was really undecided on, but still chose wrong in the end. The newscasting one felt the most difficult, due to the translucent desk which really made it difficult to trace the light/reflections.
Interestingly, a few jumped out to me as obviously real or obviously AI within just a few frames. Others I had to study intensely. It'll be interesting to study what these "tells" are (lighting, sound, compression artifacts?), and how they adapt over time.
AI trained visual clues based detection will be used to fuel better AI videos immediately.
We have now entered into a new age. What can we use now to break out? Constant video recording from each person, so that only videos with plausible continuity (verified by continuity based analysis AI) can be accepted as truth?
No, we go back to the old ways: truth is what everyone in a certain space agrees upon.
The period of modernity up to now only has this absurd concept of Truth as ultimate and self-evident thanks to limits on who had access to distributing their ideas. The emperor has his invisible suit because he's the center of power and thus defines truth. Stalin could always alter documents.
However, democratized mass media in the form of the internet has splintered the myth of a single coherent reality already, without convincing AI generated "evidence". We are reasonable creatures, not rational ones with perfect senses, and our internal perceptions have always had to be contextualized, with previous data becoming present assumptions. The reason comes after that stage. We can blame bad actors (Fox, bot farms) spreading misinformation all we like, but the fact is that there has always been a diversity of "realities", even through the peak of centralized information channels. Differences of religion, political ideals, who killed JFK, colour sensitivity, who is attractive, we call these things varying degrees of "acceptably subjective", but they're all fundamental aspects of each subject's conception of the world. This was only possible because of shared beliefs among the gatekeepers of communication.
We need to stop relying on authorities to understand things for us before we can stop relying on them to be honest. We need to stop relying on them to be honest in a world split between fascism (lite?) and Maoism (with market features), because they will never agree, and never have. We collectively avoided the crisis of the death of god solely via inertia and assuming that nobody would do such a thing as tell lies. But now nobody agrees who is the bad guy, so we can't attribute variation in worldviews to bad people.
Edit to add: I'm framing this so black-and-whitely because the only alternatives rely on either a complete revolution into utopia, basically starting tomorrow, or centralized authorities. Even the most trustworthy people and organizations can be co-opted, however. Then we would get to experience what 1984 was really about.
(also, just to be clear, I think Oswald killed Kennedy, and acted alone)
This meshes with my point of view.
It feels like humans are no further from the truth than we have been; if anything, we are closer to the truth, if such a thing exists, because we have more information.
The information is flowing so much that we can see more of the diversity of viewpoints that have always been there. More humans can see orders of magnitude more viewpoints than ever before, and the information and memes are bouncing around faster.
This could all be of great benefit to us if our nervous systems could adapt to process it properly. But who knows if we'll survive long enough to adapt into a human society that can know that much of itself.
And what rough beast, its hour come round at last,
Slouches towards Bethlehem to be born?
Thanks for bringing up the Second Coming, it's a wonderful balm for these times. One of the few that does not diminish the weight of the pain it soothes.
I'm more cynical about even hypothetical truth access myself, but it's certainly true that we have more useful models than ever before. Time and time again, people have urged those in power to democratize the responsibility of democracy, to reduce the social focus on production for its own sake, and to deescalate international tensions. The power simply continues to accrue and seems locked into its vector of motion. More and more, I see normal people catch onto this drift, and this is my source of hope. It requires buy-in. It requires us all to subjugate ourselves in the name of how things have to be, as defined by that very subjugation. Nobody can save us, but we can pull out of the dive.
I type this sincerely, directed toward anyone reading it: You are strong and able to survive the vertigo of today and the possible horrors ahead. Your love for others and the efforts it drives will be felt. I love you, and hope you love yourself enough to believe it. That is enough.
Thank you for that last thought, wervenyt, and may it also be with you. I had this thought yesterday that at various points in history, things didn't turn out as poorly as they could have, and even in the very worst of it, there are some who survive who go on to love and be loved. We'll be okay because we have each other. Hope does not come from grand changes or radical regime collapse etc, but when we connect with fellow humans one on one, or in very small handfuls, and we're kind to each other and doing small loving hopeful
Jung'sthings together.Argh, I've been hit!
Agreed. And even if we will not be okay despite that love and one another, it's the table stakes. Survival cannot be without either, that is at best death delayed by parody.
(Did you mean to write "things" where it says "Jung's"? If you meant Jung's, that might change our agreement a bit.)
Ahah foiled again by autocorrect. No, Jung's are not particularly known for their propensity for hope.
Public key infrastructure that goes down to the firmware in cameras. At least then your average scammer can’t fake a video. Big tech companies or state actors can. But it’s an improvement.
Both Android and iOS devices have most or all of the tech necessary to create attestable images.
It's a good idea, but I suspect that most people won't care that much. It might help reporters.
It’s important to have as a tool for all sorts of situations. Any time money is involved and you want to verify a photo, passports, in courts, in news. Most photos don’t need to be verifiable. But some must be.
google have synthid as a edit-resistant watermark on their AI content, and currently have given journalists access to a detector for it, but when competing models get better they're unlikely to have similar watermarks unless this gets legislated on
I'd rather make regulation to disclose when a video is generated mostly with AI. Best way to remove a gift is to make them honest.
If people still don't care from there... Well, not much surprises me in 2025.
In its current iteration Veo mostly spits out tinny sounding and low quality human voices. It's a pretty obvious tell IMO, though I'm not sure how it translates to lower quality output (e.g. a phone speaker) or how long it will remain so obvious. Ignoring visual inconsistencies, only the makeup tutorial wasn't immediately obvious from the audio to me, because certain women's voices recorded on a low quality microphone (phone) could come out sounding a bit like that.
FWIW I got 10/10 but am in no way discounting that short form AI generated video can be pretty convincing to a large audience these days, especially in the context these would normally be viewed in (uncritically, on a phone, single loop)
I got a 10/10, but I basically did it through audio only. AI sounds like it's being run through a heavy noise reduction filter, and everything has the same audio quality, regardless of how "professional" or "non-professional" the setting is. The videos that were real had reflections, background noise, and different levels of compression that the AI didn't have. I don't think I would have gotten 10/10 on mute, though.
I don't think our scores matter all that much because the point is already made: AI videos are good enough to fool lots of people. This particular test was done in such a way as to try and fool you, but it doesn't really matter. Hardly anyone will get 10/10, so it's already a scary future we are in.
I've been saying for a couple of years now that we need transparency regulations. AI absolutely must be labeled clearly as such, and it's scandalous that governments are dragging their heels on this. I know it won't stop people publishing fake videos, but it will be a start and at least stop major brands from doing it.
9/10
Spoilers
And honestly, had I looked more critically at at #7. I would have definitely gotten 10/10. The audio in #7 was particularly good, which was the most glaring issue in the other ones.Im surprised most people got the crypto bro wrong, which emphasizes how people's reasoning is biased. Only 35% guessed it correctly although the video and audio was perfect!
The false positives are the most interesting part of this imo. My theory is if someone doesn't know what to look for they fall back to what they think the apparent difficulty of faking it using AI would be. A single subject in a common scenario with nothing going on in the background? Dead simple - must be AI.
Interesting, I had a hard time with the audio in a handful of them. Two of my mistakes were falsely accusing real videos, including the one you mentioned.
The videos didn’t work for me from archive, but my library has a 3 day all access pass to New York Times digital. If you keep the page open and log in, you can refresh the original page and it should work without going through the surveys.
For some reason I thought interactives don't require a subscription. Here's a gift link for anyone who needs it.