Was just watching a YT video by part_time_Larry that said that as Microsoft own GitHub and they are a supporter of OpenAI they are using all the content on GitHub to train AI to program in any language. Probably where your example comes from
Everyone can read the public repositories on GitHub. That’s how anyone can scrape the data for machine learning. There is no need to own GitHub for this. Everyone can legally get the data, with very little effort.
It’s “open source” after all…
Mikkelsen blending allows smooth, non-repeating UV-less tiling with tiling textures. A demo:
Research paper: https://jcgt.org/published/0011/03/05/paper.pdf
Hopefully something like it can become part of Blender one day!
Game tools developer ZibraAI has unveiled Zibra VDB Compression, a new AI-based technology for compressing volumetric data in OpenVDB format “up to 20 times”.
ZibraAI claims that the tech, which will be available next year via Unity and Unreal Engine plugins, makes it possible to display smoke and fire simulations of a complexity previously only possible in VFX work in-game.
The firm also plans to release the technology as a plugin for Houdini.
Future is now old man
A few years ago image generating algorithms results were funny now there is a protest against them on ArtStation so…
This sort of thing already existed a few years ago, but instead of text to 3d, it was image to 3d:
Of course, they used some newer techniques, like diffusion to generate the point cloud, but overall, there hasn’t been that much progress when it comes to the point cloud generation.
I think the main point they were making wasn’t that it was particularly innovative in and of itself, or even as good as other solutions, but that it was much faster.
Yes, that’s exactly the point (pun was not intended ). It is worse than previous methods, but way, way faster (600x). The speed of generation was one of the things that was holding those systems back (IIRC they were able to generate a model in a day, day and a half before).
Repository, not for discussion here.
I think you should make and deserve a post of your own UuuNyaa for your tool.
It will be good to have this for rendering in Cycles
At the rate things are going, kids in the near future will just place an AI version of themselves on the screen for the teachers while they go outside to play in the snow (in the cities here in the US that completely did away with snow days because of remote learning tech.).
It is not like teachers and professors are not already having headaches with tech, since the new ChatGPT technology can create papers that earn an A while being almost impossible to catch.
Yes, replace what makes us human ! they’re probably working on facial expression support as we speak -how long till we don’t have to smile anymore ?
“Yes, it’s true : you can finally stop paying attention in all of your zoom meetings”
He says that under the pretense of a joke, but look at the technology : that’s exactly the idea.
Personally, I think it will ultimately lead to the collapse of our societies. Society is bad enough as it is with overpopulation and the welfare state. Just imagine how bad it will be when these machines lead to 90% unemployment…
That’s one of my concerns as well, but I am somewhat optimistic we will be able to figure out a new economic system to deal with that (some capitalism / socialism hybrid maybe). My major concern is information. We already somewhat do, but with deepfakes and all sorts of image- and text generating AI systems becoming better and better, we will propably live in a world in a few years in which no information can be trusted and politics will just crumble and fade away, nobody will be able to agree on the same objective reality. That’s the stuff of my nightmares, honestly.