| View previous topic :: View next topic |
| Author |
Message |
|
|
|
Sinxi and heylog
Joined: 08 May 2025
Posts: 209
|
 Posted: Sat Jan 24, 2026 9:30 am |
 |
|
|
The timing is so funny, I was just thinking "AI laws for transparency will start popping up, its a matter of when not if" and then I come to this site and I see this lol. The law will definitely need tweaking here and there, but I hope it works out well.
|
| Back to top |
|
|
|
Kougeru
Joined: 13 May 2008
Posts: 5811
|
| Posted: Sat Jan 24, 2026 10:18 am |
|
|
| Quote: | | The law does not regulate individual users who simply use AI tools, but it does apply to platforms and companies that offer AI-powered creation tools or distribute AI-generated works. |
It should apply to EVERYONE, especially when individuals are the ones using it the most nefariously to spread misinformation or blatantly fake "Evidence". But it's still hilarious how almost every country is getting better AI laws than the US
|
| Back to top |
|
|
|
ZelosZoidberg
Joined: 23 May 2018
Posts: 1076
|
| Posted: Sat Jan 24, 2026 12:03 pm |
|
|
|
I think we should limit AI to science work and ban AI for entertainment purposes. The only way I'll accept AI in entertainment is if it's sentient. For now we need people to disclose if they use AI in their work so laws like this are a good first step. I'm hoping Valve will add an AI assets tag or filter in the near future.
|
| Back to top |
|
|
|
TheRealMaria
Joined: 09 Jul 2025
Posts: 123
|
| Posted: Sun Jan 25, 2026 2:33 am |
|
|
|
As far as I'm aware most if not all AI already have watermarks on them. At least you can ask something like Grok if a picture is AI and where it was made and it'll tell you what model/site produced it. Big ugly visible watermarks on stuff is pretty lame though. Although you can easily photoshop those away like people already do with the usual visible watermarks. But we as individuals don't have to worry about it if it doesn't apply to us so it's up to the companies to deal with it.
|
| Back to top |
|
|
|
|
TdFern 87
Joined: 03 Jun 2017
Posts: 296
|
| Posted: Sun Jan 25, 2026 8:34 am |
|
|
|
If only other countries follow Korea's example.
|
| Back to top |
|
|
|
|
AsleepBySunset
Joined: 07 Sep 2022
Posts: 315
|
| Posted: Sun Jan 25, 2026 8:51 am |
|
|
| TheRealMaria wrote: | | As far as I'm aware most if not all AI already have watermarks on them. At least you can ask something like Grok if a picture is AI and where it was made and it'll tell you what model/site produced it. Big ugly visible watermarks on stuff is pretty lame though. Although you can easily photoshop those away like people already do with the usual visible watermarks. But we as individuals don't have to worry about it if it doesn't apply to us so it's up to the companies to deal with it. |
I consider AI intrinsically ugly so I don't care if they stick a big watermark all over slop, the bigger/harder to remove the better. And multiple layers of watermarking too. Cryptographic, metadata, an overlay, and also permanently saving the prompt+seed+model combo, I mean its a surefire way to know if a pictures ai, if the prompt+seed matches, it's ai.
|
| Back to top |
|
|
|
|
Purple Tentacle
Joined: 03 Aug 2025
Posts: 50
|
| Posted: Sun Jan 25, 2026 12:12 pm |
|
|
| AsleepBySunset wrote: | | I consider AI intrinsically ugly so I don't care if they stick a big watermark all over slop, the bigger/harder to remove the better. |
Considering how varied and perfect AI can be at mimicking art styles saying it's 'intrinsically' ugly doesn't really work. The quality argument really doesn't hold up anymore when people constantly mistake AI art for real screenshots or pictures these days. That's the whole reason they're proposing this law in the first place because otherwise you can't tell if something is real or fake or not.
|
| Back to top |
|
|
|
|
MelodyShaper
Joined: 01 Jul 2025
Posts: 25
|
| Posted: Sun Jan 25, 2026 6:40 pm |
|
|
| TdFern 87 wrote: | | If only other countries follow Korea's example. |
Imagine going back in time 5 years and telling people that the hate for AI would brainrot people to the point they'd unironically root for more restrictions and dictatorships.
|
| Back to top |
|
|
|
|
Egan Loo
Joined: 25 Feb 2005
Posts: 1471
|
| Posted: Sun Jan 25, 2026 7:36 pm |
|
|
| MelodyShaper wrote: | | TdFern 87 wrote: | | If only other countries follow Korea's example. |
Imagine going back in time 5 years and telling people that the hate for AI would brainrot people to the point they'd unironically root for more restrictions and dictatorships. |
Please point out where in this discussion that people have been rooting for "dictatorships."
|
| Back to top |
|
|
|
FishLion
 Crazy Fangirl
Joined: 24 Jan 2024
Posts: 861
|
| Posted: Sun Jan 25, 2026 7:36 pm |
|
|
| Purple Tentacle wrote: | | Considering how varied and perfect AI can be at mimicking art styles saying it's 'intrinsically' ugly doesn't really work. The quality argument really doesn't hold up anymore when people constantly mistake AI art for real screenshots or pictures these days. That's the whole reason they're proposing this law in the first place because otherwise you can't tell if something is real or fake or not. |
The issue with buying these products is it's great at giving an impression of quality but rarely holds water for paid products at full length. It's great at taking a screenshot from a show and altering it because the animators already did the hard work of making so many cohesive frames for it to sample, putting something else in the show and making it look cohesive for one frame isn't a huge challenge technologically. Again, it's great at stealing, it can steal every frame in an animation and produce a convincing fake but using that in a project is much trickier.
Actually inserting those frames (or comic panels for webtoons) into a cohesive whole or making it fit the rest of your project in a way that works beyond fooling people who share screenshots on social media is much harder. The law is to protect people from getting ripped off so they don't have to closely inspect every work to know if it's AI or not. There are always times people can't tell but I got an AI cookbook for Christmas and it looked convincing on the surface however the inside devolved into nonsense the further you went. Not to mention stories about people being poisoned by following mushroom foraging guides they didn't know were AI. There are a lot of times AI appears to do its job on the surface only to show cracks the longer you spend with what you paid for and people should know that what they are buying may not be checked by a human.
Even outside of media and subjective quality, there are a lot of safety and quality concerns that warrant this sort of labeling for sure, I think it's fair to say they are often low quality products that often don't do their job and fool consumers info paying for things they didn't mean to whether or not there are examples of people not knowing the difference.
|
| Back to top |
|
|
|
Greed1914
Joined: 28 Oct 2007
Posts: 5375
|
| Posted: Mon Jan 26, 2026 1:07 pm |
|
|
| Egan Loo wrote: | | MelodyShaper wrote: | | TdFern 87 wrote: | | If only other countries follow Korea's example. |
Imagine going back in time 5 years and telling people that the hate for AI would brainrot people to the point they'd unironically root for more restrictions and dictatorships. |
Please point out where in this discussion that people have been rooting for "dictatorships." |
Yeah, this sounds far more like labeling a product's country of origin. You have some way of knowing where the thing came from and can decide what to do with that information. A more informed user/customer is not a bad thing.
|
| Back to top |
|
|
|
Ashen Phoenix
Joined: 21 Jun 2006
Posts: 3041
|
| Posted: Mon Jan 26, 2026 4:40 pm |
|
|
| ZelosZoidberg wrote: | | I think we should limit AI to science work and ban AI for entertainment purposes. The only way I'll accept AI in entertainment is if it's sentient. For now we need people to disclose if they use AI in their work so laws like this are a good first step. I'm hoping Valve will add an AI assets tag or filter in the near future. |
I agree. AI has a lot of positive potential to help in the science and medical fields, but all I've seen it do for the entertainment/creative spaces is cause harm and discourse.
|
| Back to top |
|
|
|
|
