How YouTube is Tackling AI Generated Content
Вставка
- Опубліковано 8 лип 2024
- A discussion about AI generated content and how platforms such as UA-cam are trying to tackle it.
UA-cam: Disclosing use of altered or synthetic content - support.google.com/youtube/an...
Video Timeline
----------------------------
00:00 - Introduction
01:05 - Finding a new equilibrium
02:58 - AI needs original content
05:12 - The rise of AI content
06:53 - How UA-cam are tackling the deepfake problem?
10:49 - Will we see pushback?
12:14 - Final thoughts
Find out more about me at www.kevinmuldoon.com
#aicontent #deepfake #youtuber - Наука та технологія
Interesting discussion - the YT algorithm is said to simply follow what the audience likes to watch, hopefully in time the viewers will grow tired of AI content and favor content where humans show their presence ...
Unfortunately, it doesn't always work the way it should. There's always techniques that people use to chest the system.
For example, sending cheap traffic to videos 5o gain views or adding fake subscribers. It's not a major problem, but does occur.
I do think you are right though. If the audience starts rejecting cheap and easy automated content, there will be a drive for original content with real people.
PLEASE Everyone use the Feedback form (Bottom of Account menu, top right) Make it known how you feel about fake videos with fake scripts and fake voices.
Ai generated content is rife on tiktok as well, if not more so. I think there's definitely a fine line there.
It's everywhere. I actually just watched a short where a short about marvel films had a pinned comment by the uploader. In that comment, they apologised for using a different AI voice for narration. Crazy.
I need a setting to tell youtube, I DO NOT WANT AI BULLSHIT IN MY SEARCH RESULTS. I'm so sick of it.
It's everywhere now. UA-cam shorts especially
I think AI generated content is both fascinating and hilarious!
Already AI software can produce breathtaking scenes!
But sometimes you notice some disgusting bugs, artifacts of impossible IRL things on those video examples.
So I think, that the only piece we need to know, is this info true or not? Right?
Currently I think, concerning only the question at hand, that a video content author is less important than content itself (AI generated or shot by a real human on camera).
And I'm not talking about rights of ownership here (usually author should have the rights for their intellectual property).
The info on the AI video itself is the concern.
The AI video is untrusted by default.
It is hard to blocklist every possible ai permutation of the video (like a virus), but you definitely should be able to block some videos on YT if you need, like an antivirus!
So I think it is better to whitelist sources, that provide useful and real information!
Like: "Show me ai gen content only from channels I follow", or no any artificial content at all.
And if you see some on YT, you should be able to report it.
That is my current opinion on the topic.
That's a very interesting take Chuck. You've raised many interesting points.
AI is developing so quickly that I find myself changing my opinion on certain aspects of it frequently.
I do like the idea of UA-cam labelling deep fakes and AI generated videos, but not sure how that works in practice as it will be difficult to determine in some instances.
I guess the problem facing UA-cam is "Who should be ranked first?". If an AI tool used several original videos as the basis of its auto-generated video, should it be ranked higher if the video is high quality?
It's going to be interesting to see how the music and TV industries adapt to AI.
My current stance on this:
1 High quality AI video, that improves life of society, will be ranked highest, so people will share the link to it more = more ads revenue. Label it "AI", optionally provide links to original work.
2 Low quality AI video, but helpful and realistic - label it "AI". I think YT may suggest to reupload a better quality content so it will create more views, no direct actions needed.
3 Any AI video that is misleading - should be taken down immediately out of public access and blocked by YT itself. And author must face the fury of the banhammer! So it will stop fakes and useless trolling, yay :D
Cause I consider YT as a wonderful learning platform, not as some sort of a cheap fakespreading trash.
@codeChuck UA-cam have taken the first actions in acknowledging AI videos, but I believe more changes will be made in the future as AI evolves. It's clear that UA-cam recognise that this is something that isn't going away.
yes, now youtube become a garbage collector, videos teach people how to create monetizable shorts by AI automation got millions of views, this is a new gold rush. I think at the end, original content creators will leave UA-cam for a new kind of platform.
I think you could send right. I've seen a dramatic increase in AI created content this year. That means that UA-cam is pushing these AI videos to our eyes before original content.
Interesting video? I hate AI video with a passion as it’s just wasting my time. Really hope UA-cam would start flagging those videos as AI so I can avoid clicking on it.
They are taking steps to do that, but it's relying on users sharing their video is a deep fake etc.
I'm not sure if all videos that were generated by AI will have a warning/flag attached in the future.
Unsubscribed from you.
Genuinely thought you were open to honest conversation and opinions.
Saddened to experience you will delete comments expressing free speech and opinions.
Why ask questions if you are just going to delete answers you don’t like?
Shame on you!
I have no idea what you're referring to.
I only delete spam and racist comments etc. I've not even had to do that in months as UA-cam seems to be catching all the spam.
My guess is you've not successfully posted the comment or UA-cam automatically marked the comment as spam (I'll check to verify that).
I've just checked and no comments are in spam. Last comment I can see from you is the raspberry pi video, which I replied to.
I'm not sure what occurred here, but your accusations are false.