Synthetic Media & Generative AI News and Discussions

[wjfox]

[firestar464]

https://chatgpt.com/share/67058611-1a14 ... 5abd29795d

I got GPT-4o to explain AI image generators. What are your thoughts on this? I was hoping to build an FAQ on genAI for use in debates with the help of this.

Feel free to suggest topic the FAQ should cover or suggest clarifications

[raklian]

Look at the bottom right at the very end. The reporter wanted the last shot.

[Time_Traveller]

Adobe starts roll-out of AI video tools, challenging OpenAI and Meta

October 14, 20242:07 PM GMT+1

Adobe (ADBE.O), opens new tab on Monday said it has started publicly distributing an AI model that can generate video from text prompts, joining the growing field of companies trying to upend film and television production using generative artificial intelligence.

The Firefly Video Model, as the technology is called, will compete with OpenAI's Sora, which was introduced earlier this year, while TikTok owner ByteDance and Meta Platforms have also announced their video tools in recent months.

Facing much larger rivals, Adobe has staked its future on building models trained on data that it has rights to use, ensuring the output can be legally used in commercial work.

San Jose, California-based Adobe will start opening up the tool to people who have signed up for its waiting list but did not give a general release date.

While Adobe has not yet announced any customers using its video tools, it said on Monday that PepsiCo-owned (PEP.O), opens new tab Gatorade will use its image generation model for a site where customers can order custom-made bottles, and Mattel (MAT.O), opens new tab has been using Adobe tools to help design packaging for its Barbie line of dolls.

https://www.reuters.com/technology/arti ... 024-10-14/

[wjfox]

Adobe's new image rotation tool is one of the most impressive AI concepts we've seen

By Daniel John
published 2 days ago

While Adobe's annual MAX conference gives the company a chance to unveil its latest features, it also lets the brand demonstrate some of its more weird and wonderful ideas. 'Sneaks' sees Adobe engineers take to the stage to share ideas that may or may not see the light of day, such as 2024's Project Turntable. Creative Bloq enjoyed an exclusive preview of the concept in ahead of its unveiling at MAX, and it's one of the most impressive Sneaks we've seen.

Project Turntable lets users easily rotate 2D vector art in 3D, whilst ensuring it still look like 2D art from any new angle. And even after the rotation, the vector graphics stay true to the original shape, maintaining the design’s original essence.

https://www.creativebloq.com/design/ado ... -weve-seen

[wjfox]

[raklian]

It won't be too long we'll be able use VR/AR glasses to interact with realistic avatars such as this. Within 1-2 years, NPCs in video games will at least achieve this level of realism, not to mention online avatars matching what their real-life users look like.

[firestar464]

Researchers develop a new generative adversarial networks model that stabilizes training and performance

https://techxplore.com/news/2024-10-gen ... lizes.html

Good but still lags behind SOTA diffusion models on some benchmarks. IDK I'm just spitballing here

[Vakanai]

It won't be too long we'll be able use VR/AR glasses to interact with realistic avatars such as this. Within 1-2 years, NPCs in video games will at least achieve this level of realism, not to mention online avatars matching what their real-life users look like.

Almost no gamer nor person wants that last thing though. I don't want an avatar that actually looks like me.

[firestar464]

Imagine playing as yourself and then watching yourself get mauled to death by monsters or something...

[Yuli Ban]

I've mentioned going back to at least the mid 2010s that it would be AI that allowed us to achieve "true photorealism" in video games and simulations

That's not to say that the whole "40 teraflops for photorealism" prediction from a while back was wrong, only that most of those teraflops would be used for rendering the AI that allowed for said photorealism, not necessarily rendering polygons and lighting.

[raklian]

Imagine playing as yourself and then watching yourself get mauled to death by monsters or something...

That would actually be cool.

[Tadasuke]

Almost no gamer nor person wants that last thing though. I don't want an avatar that actually looks like me.

I've always thought that video games can be great for people who are dissatisfied or bored with reality, including their own bodies. Looking as yourself can be perhaps useful in some business, scientific or classroom meetings. Otherwise not so much.

If you love how reality works then go have adventures in reality. Don't play video games. If you don't like, then they are perhaps the main alternative (because of some level of interactivity).

[Tadasuke]

It's really saddening and disheartening, seeing these kind of messages popping more and more often:



If you asked me in 2011, I would had told you that in the 2020s humans would somehow upgrade their brains to be better artists and make ambitious art quicker. Not this...

Back in 2010-2013 I made a list of "100 awesome ideas for advancements of the next 20 years". Nothing has come true as of yet, but meanwhile many bad, dumb or questionable things have managed to happen (as usual, because of human pride, recklessness and inconsideration).

Imagine following someone (who has worked hard) for years only to see them delete everything and stop uploading because of some weirdos generating countless AI-made crap ... it makes you question everything.

[raklian]

[firestar464]

[Vakanai]

It's really saddening and disheartening, seeing these kind of messages popping more and more often:



If you asked me in 2011, I would had told you that in the 2020s humans would somehow upgrade their brains to be better artists and make ambitious art quicker. Not this...

Back in 2010-2013 I made a list of "100 awesome ideas for advancements of the next 20 years". Nothing has come true as of yet, but meanwhile many bad, dumb or questionable things have managed to happen (as usual, because of human pride, recklessness and inconsideration).

Imagine following someone (who has worked hard) for years only to see them delete everything and stop uploading because of some weirdos generating countless AI-made crap ... it makes you question everything.

Agreed. Instead of enabling artists, they're art is being taken from them without payment to train AI to make "art" in their styles, effectively to replace them. These artists wish to make a living from their art, and instead their art is being used to train something to make their livelihoods harder. It'd be one thing if we didn't live in a capitalist "work to live" society, but right now it's hard to justify or see the value of this in today's unfair society.

It might be different if the training data was only used to make AI smarter, and wasn't used for the "art" generation programs. I'd be fine with art training AI if it wasn't used to compete directly against these same artists, or the artists were at least compensated for their art being used, or being able to opt out without having to delete everything.

[firestar464]

Basic income program for artists is the way IMO.

[firestar464]

Simplifying, stabilizing, and scaling continuous-time consistency models

https://openai.com/index/simplifying-st ... cy-models/

[Powers]

Basic income program for artists is the way IMO.

I don't know. I'm starting to believe this will NOT resolve the gap in this still scarcity society.