I just looked for ai slop on the internet and some videos are very realistic, so realistic it makes me doubt my ability to discern what’s real from what’s created to play with my emotions and generate money for the creator or advance an agenda.
I’m in my 40s. Video sites are full of what I assume overconfident people younger than me with comments like how boomers would believe any of these videos. I’m not that old myself yet but this stuff is scary. It can be used to denigrate a politician, to demonize or ridicule minorities, to share misinformation, to make porn using the face of somebody who rejected a disgruntled man…
It’s also very sad society actually wants this. It shows lots of people are actually very gullible and stupid.
A better question would be, how do I avoid being gullible with images and video so realistic? Because the more technology advances the worse it’s going to get.
Start with the assumption the video is fake. Do you trust the source? If the video is of something outrageous, did you notice something feels uncanny? Does the lighting, shot angle, and proportions make sense?
If the answer is not yes, no, yes then it’s probably AI
Alot of voices are still pretty bad, like words and voice sound good, but their tone and emphasis is off for what they are saying.
I would add that, when you listen to an AI conversation, there are no pauses. Like they know exactly what to say before the other person is done talking. Most humans have a natural pause to think about what was just asked.
Subtle unnatural movements. Everything looks too nice and perfect. Uncanny facial expressions.
For animal videos: everytime there are videos of animals acting too human, mostly in terms of showing human-like affection, its AI.
Situations that are too fantastical to be real. An example could be a video I saw of a guy in a ski lift and a big snow owl lands on his shoulder.
In general: objects or subjects lacking realistic weight and/or physicality to them.
Noticing inconsistencies in the background where objects or subjects melt together, like a person in the background merging with a chair or a bookshelf for a second.
Also, different videos with very specific scenarios like for example surveilance footage of wild animals running through screen doors and into a house while a human is sleeping on the couch or different videos of go pro footage of people in Rollercoasters that go off the rails and fly into the air.
Things like these, I have noticed.
The tells are getting subtler but currently a good tell is the camera work. If the video looks amateurish (grainy, blurry), but the video is more stable than a $60k gimbal mounted movie grade model with active scene stabilization, it’s probably AI. Another giveaway is motion morphing items from hammerspace where there’s a visible “seam” where the LLM is trying to reconcile two prompts. (example: you see a video of a horse rider, and the bridle and reins might suddenly decrease in length instead of becoming taught when the rider wishes the horse to stop, or you may see objects morph into existence from other structures, but so fast you might have to rewatch it a few times to catch it.)
It’s worth pointing out that whatever tips you get here, accurate though they may be today, are likely to be less accurate with time and this tech is improving at a rapid pace so if this is important to you, you’ll have to keep on top of what techniques for discernment are current at any given time or if there even remain any reliable techniques at all.
Looks like eventually it’s all going to come down to context and your assessment of the trustworthiness of the source. If something is so sensational it seems unbelievable and you don’t have much trust in the source from which you saw it, it probably is literally unbelievable until further corroboration emerges. This is a pretty exhausting model when modern live involves so much media consumption but I’d advise that there’s also an element of practicality you should incorporate in to your evaluation. You’ll have to decide if the answer to the question “is this AI?” is important for any given situation before actually investing any time or energy in to answering it. Sometimes you might get the calculation wrong and it turned out that by assuming something was innocuous enough not to matter one way or the other you ended up being misinformed about something that actually was important but at the end of the day we’re only mere mortals and can sometimes be wrong.
Whenever I don’t like the video or the person in it it’s ai.
It’s getting harder and harder but some things still seem to work:
- emotional emphasis on the wrong syllables
- dissonance between facial expressions and vocal delivery
- AI writing giveaways (it’s not A it’s B etc.)
- dreamlike / “floaty” motion
- unrealistic “depth of field” (objects don’t blur properly with distance)
- unrealistic lighting / coloring / appears stylized despite the attempted hyperrealism
- object permanence problems / subtle drift in sizes and proportions over time
The last one is the big giveaway but it can be hard to spot. Look for things going out of frame or going behind other objects. Do they come back when they should? Have they changed?
For example: there are three people in the background and a bus drives by, obscuring them completely. Are they still there when the bus moves on?
I’ll add another one: multiple vanishing points. As models will pull different parts of images from different sources, perspective can be inconsistent across the image.
I would be concerned with the type of media I’m consuming in the first place if ai can make it problematic
The same way you know if there’s the Fae.
- Count the fingers
- Look for irregularities
- See where the glamour shimmers
- Ask them for their name
Lettering can give it away also.
GenAI doesn’t have problems with fingers anymore, as far as I know.
It does, but less often
Perfect framing is often a giveaway. Most people’s phone footage wobbles and moves around. Background object permanence is another thing to look out for
The Corridor Crew have done a couple of videos where they explain a lot of the things to look for and the reason AI finds it hard. Some examples:
Was going to recommend the same thing, but you did it better (with links).
Something else I remember them pointing out is that a lot of the digital video generators can only generate around 10-15 seconds at a time, so videos that are that long are more suspect. Longer videos can be done, but they basically have to feed the second segment of video from the last frame of the first, but the AI doesn’t have context for what it is and that can cause things to drastically change what they are doing or how they act at around that 10-15 second mark. (like a ball that was previously bouncing is now just rolling like all the momentum was stolen by magic). It isn’t a magic bullet (there unfortunately isn’t one) and as the generators evolve it might disappear, but it has helped me in a few cases. And it feels like once you have one good piece of evidence that something is AI generated, you start to see others popping up all over.
Was about to recommend Corridor’s videos! They are really comprehensive and easy to follow.
A better question would be, how do I avoid being gullible with images and video so realistic?
Honestly, we’re quickly moving into “assume AI unless proven otherwise” territory.
Don’t assume that recognizing fakes is a younger generation thing. Maybe the people spotting them grew up reading snopes and alt.folklore.urban so that they have a good idea of how urban legends, hoaxes and clickbait are constructed.
Checking the channel that uploaded can also give a clue. A lot of the slop channels just flood variations of the same thing. They just post a thousand generated videos of different dog doing the exact same thing or “people” having the exact same fight etc.
Sorting videos by oldest can help too, if their first video predates AI it’s probably a decent indicator the channel is real
I wouldn’t bother learning, whatever you learn today could be out of date in a year and could overall be impossible in less than 5 years. historically people would believe whatever title you slap on a youtube video regardless of if the video even tried to demonstrate the claim or not. I’d focus on authoritative and reliable sources and just be an open minded but also vigilant skeptic of claims you have no practical way to validate.
for example they’ll show you something like a video of Latino Americans celebrating with overlay text claim it’s Venezuelans supporting Trump (I’ll assume you don’t know Spanish or there’s no original audio) and it’s a real video but the original video was celebrating a sports game like 10 years ago.
That guys voice.
Plus, minor grammatical errors. eg. “Has increased the Tens of times”
That woman’s voice too.
And it’s not just Bruce Lee having a fight.
It’s angry conceited Martial Arts fighter picks random person out of a crowd for a demonstration,
who happened to be none other than calm collected Bruce Lee, yet no one knew who he was,
with 4k grayscale picture quality and at all angles of the dojo, including on individual audience members themselves, plus repetitive ‘damaged film’ effects for 20th century authenticity.
