YouTube Question Tone of voice, Yelling, not just swearing, can apparently get you age restricted.

[Beanie Draws]

So, a creator I follow has been doing some tests after their content constantly gets age restricted, and he doesn't swear in his content, but he IS known for yelling intensely in his intros. There's no "anger" in his intros, it's simply an intensity, much like a drill sergeant, and apparently it's that, that has got his content age restricted.

He did a test doing a calm introduction, and the video DIDN'T get age restricted.

This is both interesting and concerning. I know the AI scans every frame of your video for as much context as it can, and also scans the audio for copyright and swearing (it's also how it can detect auto captions) but it seems it also judges the volume and tone of voice, and if it interprets your video as "angry" it will age restrict your content, even if your content is fine for all audiences.

I'd be interested to know if anyone else with a more intense way of speaking, has encountered similar situations? I speak in a calm tone of voice so I've not noticed any age restricting of my videos, but I can see how "angry" voices can get you age restricted, even if the video it'self isn't angry nor contains any profanities.

 

[tropicthunder]

Interesting. I just know about this. Is this new AI feature from youtube or it's been there for quite some time ?

Auto caption most of the time incorrectly translate my language (Indonesian) into English, but it's not a big deal for me. Also sometimes, they're translating a caption that doesn't exist, by not exist I mean there's no dialogue.

 

[Yly66]

Very interesting subject, I didn't know that it restricts the videos for the tone of voice

 

[Drscottbland]

So, a creator I follow has been doing some tests after their content constantly gets age restricted, and he doesn't swear in his content, but he IS known for yelling intensely in his intros. There's no "anger" in his intros, it's simply an intensity, much like a drill sergeant, and apparently it's that, that has got his content age restricted.

He did a test doing a calm introduction, and the video DIDN'T get age restricted.

This is both interesting and concerning. I know the AI scans every frame of your video for as much context as it can, and also scans the audio for copyright and swearing (it's also how it can detect auto captions) but it seems it also judges the volume and tone of voice, and if it interprets your video as "angry" it will age restrict your content, even if your content is fine for all audiences.

I'd be interested to know if anyone else with a more intense way of speaking, has encountered similar situations? I speak in a calm tone of voice so I've not noticed any age restricting of my videos, but I can see how "angry" voices can get you age restricted, even if the video it'self isn't angry nor contains any profanities.

I was disappointed because I like the ΓÇ£man about techΓÇ¥ and that was his thing

 

[Stanley | Team TB]

Wow... I have heard some weird things about Google Brain (the AI they use for this sort of stuff). I never heard anything about yelling, but it makes perfect sense in conjunction with the other stuff they use t for.

 

[Stanley | Team TB]

I was disappointed because I like the ΓÇ£man about techΓÇ¥ and that was his thing

You have great taste!

 

[HenryBR]

Very interesting thread. Who would have known that tonality in one's video can have such an affect. Thanks for the heads up!

 

[Beanie Draws]

Interesting. I just know about this. Is this new AI feature from youtube or it's been there for quite some time ?

Auto caption most of the time incorrectly translate my language (Indonesian) into English, but it's not a big deal for me. Also sometimes, they're translating a caption that doesn't exist, by not exist I mean there's no dialogue.

I believe the AI has been there since around 2017/2018

It was actually Roberto Blake that highlighted a deeper issue with the AI situation. An article about it is here https://blog.cloudhelix.io/the-truth-about-youtubes-monumental-ad-crisis-7430d00fc255

Essentially, it's nearly impossible to moderate YouTube content for genuinely harmful material because of the sheer number of uploads every minute, of every hour, of every day, so they've had to implement AI. That article above was about another AI creator was suggesting YouTube couldn't do it effectively without violating their patent. Very interesting story.

But this has been a growing issue for years, getting even more intense in 2017, and it seems some innocent creators get caught in the crossfire along the way. You can get your content manually reviewed and taken out of restricted mode, but essentially, all the errors you see in your auto captions that YouTube provides, those same errors can happen with the context of the way you speak as well.

 

[dolewhipdad]

I am interested to learn more about this! I can't imagine something getting age-restricted for tone...but who kn ows?!?! Ill be following this thread!

 

[tropicthunder]

I believe the AI has been there since around 2017/2018

It was actually Roberto Blake that highlighted a deeper issue with the AI situation. An article about it is here https://blog.cloudhelix.io/the-truth-about-youtubes-monumental-ad-crisis-7430d00fc255

Essentially, it's nearly impossible to moderate YouTube content for genuinely harmful material because of the sheer number of uploads every minute, of every hour, of every day, so they've had to implement AI. That article above was about another AI creator was suggesting YouTube couldn't do it effectively without violating their patent. Very interesting story.

But this has been a growing issue for years, getting even more intense in 2017, and it seems some innocent creators get caught in the crossfire along the way. You can get your content manually reviewed and taken out of restricted mode, but essentially, all the errors you see in your auto captions that YouTube provides, those same errors can happen with the context of the way you speak as well.

Ahh now I understand about this problem. From youtube perspective, I understand that they cant review manually each and every videos because the sheer number of it.

The bad thing is that just like you said, innocent youtuber got caught in between, so they have to appeal for the mistake from the AI, and this is one of the annoying things for youtuber I guess.

 

[Damon]

Wow!! Man, oh, man.

 

[lifeinplastic]

Following this thread. I am usually very calm in my videos, but this is interesting. I have a creator friend whose energy level is 9 or 10 on a scale of 10. I'm going to ask her if she has noticed this in her videos.

 

[DIMTIPS]

All of this makes you think of what else the algorithm is using as matrix to determine other video features.