Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

YouTube now requires to label their realistic-looking videos made using AI

blog.google

Post reply on HN
Print view View original

381–390 of 517 posts

Re: YouTube now requires to label their realistic-looking videos made using AI

#381

Post by sverhagen »

JohnFen wrote: post #239
jjcm wrote: post #45

Earlier quoted context omitted.

One black hat thing I'm curious about though is whether or not this tag can be weaponized. If I upload a real event and tag it as AI, will it reduce user trust that the real event ever happened?

I fear that we're barrelling fast toward a future when nobody can trust anything at all anymore, label or not.

I was immediately thinking that the #AI labels are going to give people a false sense of trust, so that when someone posts a good-enough fake without the #AI label, it can do damage if it goes viral before it gets taken down for the mislabeling. (Kudos for the effort, though, YouTube.)

Re: YouTube now requires to label their realistic-looking videos made using AI

#382

Post by vkou »

ravenstine wrote: post #281
vkou wrote: post #18

Earlier quoted context omitted.

Society is simply revisiting a conversation about doctored photographs, videos, and audio recordings. The last word on this subject was not written in the 1920s, it's good to revisit old assumptions every century or so, when new forms of media and media manipulation become developed. The first pass on it is unlikely to be the best, or even the last one.

> Society is simply revisiting a conversation about doctored photographs, videos, and audio recordings. society in this case = media companies

So go advocate. Go write books, op-eds. Go present at conferences. Universities. Go lobby. Rabble-rouse.

You have to speak to get heard.

Re: YouTube now requires to label their realistic-looking videos made using AI

#383

Post by albert_e »

jjcm wrote: post #37

I think it's smart to start trying things here. This has infinite flaws with it, but from a business and learnings standpoint it's a step toward the right direction. Over time we're going to both learn and decide what is and isn't important to designate as "AI" - Google's approach here at least breaks this into rules of what "AI" things are important to label: • Makes a real person appear to say or do something they…

And some of these questions apply to more traditional video editing techniques also.

If someone says "I am not a crook" and you edit out the "not", do you need to label it.

What if it is done for parody.

What if the edit is more subtle - where a 1 hour interview is edited into 10 minute excerpts.

Mislabeled videos for propaganda.

Or simply date or place incorrectly stated.

Dramatic recreations as often done in documentaries.

Etc

Re: YouTube now requires to label their realistic-looking videos made using AI

#384

Post by BlueGh0st »

summerlight wrote: post #34

Looks like there is a huge grea area that they need to figure out in practice. From https://support.google.com/youtube/answer/14328491# : Examples of content creators don’t have to disclose: * Someone riding a unicorn through a fantastical world * Green screen used to depict someone floating in space * Color adjustment or lighting filters * Special effects filters, like adding background blur or vintage effects * Pro…

>Showing a realistic depiction of a tornado or other weather events moving toward a real city that didn’t actually happen

A bit funny considering a realistic warning and "live" radar map of an impending, major, natural disaster occurring in your city apparently doesn't violate their ad policy on YouTube. Probably the only time an ad gave me a genuine fright.

Re: YouTube now requires to label their realistic-looking videos made using AI

#385

Post by pojzon »

Gigachad wrote: post #376
JohnFen wrote: post #239

Earlier quoted context omitted.

I fear that we're barrelling fast toward a future when nobody can trust anything at all anymore, label or not.

It just goes back to trusting the source. If 5 media orgs post different recordings of the same political speech, you can be reasonably sure it actually happened, or at least several orders of magnitude more sure than if it's one blurry video from a no name account.

And then you learn all of those media orgs are owned by the same billionare.

There will be no way to say something is true beside seeing it with own eyes.

Re: YouTube now requires to label their realistic-looking videos made using AI

#386

Post by mjevans »

jjcm wrote: post #37

I think it's smart to start trying things here. This has infinite flaws with it, but from a business and learnings standpoint it's a step toward the right direction. Over time we're going to both learn and decide what is and isn't important to designate as "AI" - Google's approach here at least breaks this into rules of what "AI" things are important to label: • Makes a real person appear to say or do something they…

Truth in 'advertising' is important.

>> 17th angle - AI Generated vantage point based on existing 16 videos (reference links). Would be • Alters footage of a real event or place

Re: YouTube now requires to label their realistic-looking videos made using AI

#388

Post by chii »

lysp wrote: post #367
jjcm wrote: post #37

I think it's smart to start trying things here. This has infinite flaws with it, but from a business and learnings standpoint it's a step toward the right direction. Over time we're going to both learn and decide what is and isn't important to designate as "AI" - Google's approach here at least breaks this into rules of what "AI" things are important to label: • Makes a real person appear to say or do something they…

I think it also gives them a legal / business remedy if people fail to label their content. If someone for example makes a political video and fails to label it, they can delete the video/terminate the account for a breach of service.

but this gives room to also abuse the uncertainty to censor anyone without recourse - by arguing such and such video is "AI" (true or not), they have a plausiblely deniable reason to remove a video.

Power is power - can be used for good or bad. This labelling is a form of power.

Re: YouTube now requires to label their realistic-looking videos made using AI

#389

Post by GauntletWizard »

sverhagen wrote: post #380
lysp wrote: post #367

Earlier quoted context omitted.

I think it also gives them a legal / business remedy if people fail to label their content. If someone for example makes a political video and fails to label it, they can delete the video/terminate the account for a breach of service.

Given the regular stories posted on HN about folks who've had some aspect of their social or other media canceled by any some SaaS company, are these companies having many (legal) qualms as it is about canceling people without providing a good reason for it? Would be nice if they did, though...

I'd much prefer Google cancel capriciously with solid TOS backing to it than without, but I'll complain about their double standards about what they choose to censor... Not regardless, but without a doubt, because Google will choose to selectively enforce this rule.

Re: YouTube now requires to label their realistic-looking videos made using AI

#390

Post by chii »

dataflow wrote: post #363
8note wrote: post #360

Earlier quoted context omitted.

If you do a jpeg compression, or crop the file, then does that signature matter anymore?

Not if you do it, only if the chip also gives you a signed JPEG. Cropping and other simple transformations aren't an issue, though, since you could just specify them in unsigned metadata, and people would be able to inspect what they're doing. Either way, just having a signed image from the sensor ought to be adequate for any case where the authenticity is more important than anesthetics. You share both the processed…

> You share both the processed version and the original, as proof that there's no misleading alteration

so you cannot share the original if you intend to black out something from the original that you don't want revealed (e.g., a face or name or something).

The way you specced out how a signed jpeg works means the raw data _must_ remain visible. There's gonna be unintended consequences from such a system.

And it aint even that trustworthy - the signing key could potentially be stolen or coerced out, and fakes made. It's not a rock-solid proof - my benchmark for proof needs to be on par with blockchains'.

Post reply on HN