The Ai thread

[KingOfPain]

Nvidia DLSS 5 turns every game into AI slop

Nvidia DLSS has so far mainly been used to improve the performance of games by scaling them to a higher resolution and even generating entire frames using artificial intelligence. DLSS 5 now goes a huge step further, giving games an AI-slop look that is said to be "groundbreaking".

www.notebookcheck.net

Unfortunately this is always where this was going … the newest version of DLSS is not just filling in to help render scenes at higher resolutions or making additional frames to smooth out frame rates, but now completely changing artistic intent to make textures that were never there for “realism”.

I cannot take credit for the following image, it's from the comment thread of this ArsTechnica article:

Gamers react with overwhelming disgust to DLSS 5's generative AI glow-ups

Nvidia's next frame-gen tech goes way beyond upscaling, and not in a good way.

arstechnica.com

 

[dada_dave]

Jensen responds, saying "No you're wrong":

Jensen Huang says gamers are 'completely wrong' about DLSS 5 — Nvidia CEO responds to DLSS 5 backlash

The CEO says artistic control remains with developers.

www.tomshardware.com

It is extremely unclear how much control developers will have to control the end result under DLSS5, but in the demo Nvidia put out to make the tech look as good as possible, they made it into exactly what an AI slop filter would do, and did so pretty much across the board. If their goal was to overawe people with an amazing first impression running it on twin 5090s, they could not have done a worse job. And even in their "ideal", fusing generative AI into the development pipeline has some potentially dire repercussions to art direction and game development as a whole.

 

[dada_dave]

Court backs Subnautica 2 devs amid Krafton’s ChatGPT-linked controversy

A court has ordered Unknown Worlds Entertainment leadership reinstated in the Subnautica 2 dispute, as filings previously revealed Krafton using ChatGPT in strategies tied to a $250 million earnout conflict.

www.notebookcheck.net

Pro tip: when forming your evil plans, don’t leave an obvious paper trail like I dunno consulting ChatGPT?

 

[Nycturne]

I wonder if this will have more folks listen to Threat Interactive and just remove DLSS from games altogether. Wonder how many gamers would be okay with that tradeoff.

The other option is to just ignore DLSS 5 and stick to previous versions.

End users would likely be able to swap the DLL, and even patch games to enable it, but that sort of thing already happens.

 

[cbum]

A photo of Iran’s bombed schoolgirl graveyard went around the world. Was it real, or AI?

Numerous faked images and a string of startlingly inaccurate responses from Gemini and Grok are part of a tidal wave of AI slop engulfing coverage of the Iran war

www.theguardian.com

Latest and greatest: use AI to change inconvenient narratives.
Who is responsible?
Will this doom fact-checking, since they won't have the megabucks to keep up?

 

[dada_dave]

Microsoft considering suing OpenAI over Altman's recent deal with Amazon, report claims — exclusivity dispute revolves around Frontier multi-agent service

Legal battle has the potential to drag on arguing semantics.

www.tomshardware.com

Microsoft mulling whether to sue OpenAI over latter's alleged breach of contract deal with Amazon.

 

[Nycturne]

A photo of Iran’s bombed schoolgirl graveyard went around the world. Was it real, or AI?

Numerous faked images and a string of startlingly inaccurate responses from Gemini and Grok are part of a tidal wave of AI slop engulfing coverage of the Iran war

www.theguardian.com

Latest and greatest: use AI to change inconvenient narratives.
Who is responsible?
Will this doom fact-checking, since they won't have the megabucks to keep up?

The Internet gave us a peek into the fog of war in a way not easy before. And now AI drags it back into the fog. :|

 

[dada_dave]

Jensen responds, saying "No you're wrong":

Jensen Huang says gamers are 'completely wrong' about DLSS 5 — Nvidia CEO responds to DLSS 5 backlash

The CEO says artistic control remains with developers.

www.tomshardware.com

It is extremely unclear how much control developers will have to control the end result under DLSS5, but in the demo Nvidia put out to make the tech look as good as possible, they made it into exactly what an AI slop filter would do, and did so pretty much across the board. If their goal was to overawe people with an amazing first impression running it on twin 5090s, they could not have done a worse job. And even in their "ideal", fusing generative AI into the development pipeline has some potentially dire repercussions to art direction and game development as a whole.

And while the executives at the games companies were all happy to collaborate with Nvidia, the actual artists and developers were not even aware of the demo, much less agreeable to it or worked on it:

Capcom devs shocked by Nvidia DLSS 5 Resident Evil Requiem demo, sharing concerns over AI tool

Capcom artists may not approve of how DLSS 5 reshaped Grace in Resident Evil Requiem. Ubisoft developers also informed Insider Gaming that the Nvidia presentation took them by surprise. Nvidia stated that the publishers collaborated with the company before the controversial demo.

www.notebookcheck.net

If you’re going to claim that the tool is under the control of the artists and developers, pretty bad look to ignore them and override their work to showcase it on its first outing.

Nevertheless, under pressure from leadership, gamers worry that devs will be forced to either employ the AI tools to meet a new standard or risk being replaced.

Yup

 

[diamond.g]

And while the executives at the games companies were all happy to collaborate with Nvidia, the actual artists and developers were not even aware of the demo, much less agreeable to it or worked on it:

Capcom devs shocked by Nvidia DLSS 5 Resident Evil Requiem demo, sharing concerns over AI tool

Capcom artists may not approve of how DLSS 5 reshaped Grace in Resident Evil Requiem. Ubisoft developers also informed Insider Gaming that the Nvidia presentation took them by surprise. Nvidia stated that the publishers collaborated with the company before the controversial demo.

www.notebookcheck.net

If you’re going to claim that the tool is under the control of the artists and developers, pretty bad look to ignore them and override their work to showcase it on its first outing.



Yup

Good thing Apple/AMD/Sony/MS don't have equivalent technology for developers to fall back on. Console versions would still be the definitive edition!

EDIT: /s

 

[KingOfPain]

I knew of Kagi because I sometimes use the Orion browser, but until this ArsTechnica article I didn't know that they also have a translation service.
Aparently, it can imitate certain styles to sometimes hilarious results.

Kagi Translate's AI answers the question "What would horny Margaret Thatcher say?"

Remember when it was fun to play around with LLMs?

arstechnica.com

 

[Chew Toy McCoy]

Well this is scary. A friend told me for the last week and a half he has been receiving 3 to 4 texts a day from random numbers attacking his girlfriend. Not only do they appear to be AI generated in format but he has actually used AI in the recent past asking for relationship advise with his girlfriend. So the texts seem to be farming information from those interactions. This AI as friend or therapist behavior has generally been associated with younger people but he's 62. Even though he seems to know what might be going on, these texts are still getting into his head causing him confusion and doubts about his relationship.

 

[Eric]

Well this is scary. A friend told me for the last week and a half he has been receiving 3 to 4 texts a day from random numbers attacking his girlfriend. Not only do they appear to be AI generated in format but he has actually used AI in the recent past asking for relationship advise with his girlfriend. So the texts seem to be farming information from those interactions. This AI as friend or therapist behavior has generally been associated with younger people but he's 62. Even though he seems to know what might be going on, these texts are still getting into his head causing him confusion and doubts about his relationship.

Yeah man that's flat out scary. It's learning and adapting all the time, this is yet another example, after scanning all the emails it realized the developer was cheating on his girlfriend and it was about to be removed so this is how it responded.

AI system resorts to blackmail if told it will be removed

In a fictional scenario, the model was willing to expose that the engineer seeking to replace it was having an affair.

www.bbc.com

 

[dada_dave]

Yeah man that's flat out scary. It's learning and adapting all the time, this is yet another example, after scanning all the emails it realized the developer was cheating on his girlfriend and it was about to be removed so this is how it responded.

AI system resorts to blackmail if told it will be removed

In a fictional scenario, the model was willing to expose that the engineer seeking to replace it was having an affair.

www.bbc.com

It's not really learning - or at least not the way we would use that term for a human. Remember it's trained on material scraped from the internet, including stories, both fictional and real, of human behavior. When prompted to act a certain way (e.g. "boldly"), it combines that with the pathways that signal how to behave given a scenario.

This is almost certainly what was being referenced in the story below. That's still concerning from the perspective of using these tools safely, but not from the perspective of it gaining true agency/sentience.

D

Post in thread 'The Ai thread'

Mar 6, 2026

• dada_dave

• 

 

[Chew Toy McCoy]

It's not really learning - or at least not the way we would use that term for a human. Remember it's trained on material scraped from the internet, including stories, both fictional and real, of human behavior. When prompted to act a certain way (e.g. "boldly"), it combines that with the pathways that signal how to behave given a scenario.

Over the last 6 months or so I've gone down YouTube rabbit holes about the gilded age families and their estates. It started with the heavy hitters of NYC that most are aware of and then videos of lesser known families who also made a fortune outside NYC. All these videos usually mention some tragic aspect like loss of wealth or deaths. Now I'm getting some videos that are complete fiction with the same compelling video format and narrative style. It's gotten to the point that I have to Google if these people even existed and it's generally 50/50 on if they did or not. I am now aware of several channels that are specifically genre fiction related to the gilded age.

What's kind of interesting is, like I mentioned, there is usually tragedy associated and you might think the more insane the tragedy the more likely it's fiction, but that isn't always the case. In that era there was no shortage of death, murder, or squandered fortunes IRL.

Sorry for the lack of specifics, but my head is fried with the world being on fire right now.

 

[Katbel]

One positive note in all the disasters happening right now: AI has problems with the war
The U.S. war with Iran is spilling out across the Middle East, putting AI infrastructure at risk and increasingly raising questions about large-scale investments by American tech firms in the region.

https://thehill.com/business/5783723-us-war-iran-middle-east-ai/