With so much AI slop on YouTube, it's no longer safe to give videos the benefit of the doubt. Always check

[highplainsdem]

to see if they're AI or likely to be AI.

We've had at least three DUers tricked by AI slop videos today. The first slop video I saw came from a channel posting only AI slop videos about earthquakes (only 5 as of this morning, but the channel had just started posting them on the 1st of July). Then another DUer was tricked by an AI slop channel posting nothing but fake Bill Clinton statements, using an AI clone of his voice - garbage that AI user had written for that fraud. 60 such videos since April 1. Then a third DUer was tricked by AI slop supposedly giving inside information from the Vatican - 232 videos about that since January 22.

ALWAYS, to avoid AI slop, check the date a video was uploaded. If it's more than a few years old, you don't have to worry that it was generated with AI and probably filled with AI errors that no one bothered to catch, if not deliberate fraud.

But if it was posted in the last few years, more checking is necessary.

GO TO THE YOUTUBE CHANNEL it's from and look at the other videos. If it's a recent video and the channel is posting videos nearly every day and sometimes multiple times a day, the odds are it's AI slop, UNLESS it's from a real news organization with a large staff.

Check the information in the channel profile and individual video descriptions. Sometimes AI users admit they're using AI, but not all do.

Check the YouTube comments on videos, because you'll often find comments saying that AI slop videos are AI.

If there is ANY use of AI art, even for the thumbnail (top image), the odds are that isn't the only such use of AI. I hope you can recognize typical AI art, which often has a finish I've sometimes seen described as looking like plastic.

If the video includes images or short video snippets that don't go with the narration, it was likely patched together by AI.

If a series of videos on the same theme have different voices narrating different videos, the odds are it's AI narration.

There is an incredible amount of AI slop on YouTube already, some of it produced by content farms or individuals that are likely to have set up other YouTube channels on different subjects, all of them just clickbait.

See this thread in Musicians, the OP and replies, for the sort of things you might run into with AI slop: https://www.democraticunderground.com/10358002

It's an AI slop swamp out there. But if you do some checking before copying videos to DU, you can help keep it from spreading here.

[Arthur_Frain]

1. Too much to hope for

But there should be an “AI Slop” channel that I can trash.

[highplainsdem]

2. On YouTube, you mean? They could do a much better job labeling AI, and letting people block it, but

they don't because they make money off that garbage.

Zuckerberg is encouraging AI slop on Facebook and Instagram for the same reason.

[Arthur_Frain]

10. No, here on DU

Used to be a “videos” tab here on DU when I first got here. Even then it was on its way out, and self policing here was non existent, as the web became more video driven.

Now the web is AI video driven. Nobody here seems interested in self policing”censoring”, everybody just wants views and clicks.

And so we’ll yell back and forth about “AI slop”.while it becomes the main posting medium here.

[moondust]

3. Thank you for the tips!

Bookmarked!

[Crunchy Frog]

4. And YouTube keeps asking me to get a paid membership. Not a chance.

I try to keep my video viewing to channels that I'm subscribed to or have been following long term and know are trustworthy. There's lots of stuff that just pops up though, and usually has clickbaity titles.

[canetoad]

5. As well as reading the blurb about the video

Which SHOULD say at the bottom if it's fully or party AI created, it can be useful to visit the channel home page and scan what is on offer.

For example - the recent post by OAITW about Pope Leo from the channel Tales of Peace - this is their video page:
https://www.youtube.com/@TalesofPeace-s9t

A quick glance should set your spidey senses tingling. This is obviously not genuine or truthful.

[Aussie105]

6. I've never taken anything on the internet, YouTube or Tik Tok seriously.

And I'm not about to do so now.

DU is the only place that should be taken seriously - as long as you can sort the wheat from the chaff.

[orthoclad]

7. I'd like to see ALL media label AI

That would make it easier to track down the stolen property.

[orthoclad]

8. Googletube is just another brick in the wall

of the googly empire. I've been avoiding it for years. Damn straight gaggle is spying on the users, for a variety of reasons, mainly commercial, but increasingly political.

Big Google Is Watching! With the processing power of its data surveillance farms, Federal cases will be built against protestors by linking statements in gmail and gchat to watching certain (possibly bait) videos on the tube.

Plus a host of other reasons to avoid videos which I sum up as "life is too short for utube". I can scan a block of text in seconds for a gist rather than waste half an hour listening to some geek blather about what he's going to show you in just a minute but first these ads... Complete with garish yellow circles lines and arrows (thank you Arlo)

In some very rare cases it's worth watching video.

[DallasNE]

11. The AI Slop

Is what makes me ask: what is so wonderful about AI? Is it really ready for prime time? I have my doubts about this and the security aspects of AI, as well as whether it can be trusted with sensitive data. As a retired IT mainframe computer programmer, I first came in contact with subsets of AI in the mid-1990s, and it has taken this long (30 years) to move forward from those early efforts.

[C Moon]

12. FaceBook "Reels" is the worst for AI videos--especially the animal ones. I would guess 90% are AI.