News for hackers. Stuff that matters.
Re:Diffusion Models Are Real-Time Game Engines by TeMPOraL on Wednesday August 28, 2024 @09:50AM (#41379524)
Like many people in case of LLMs, you're just demonstrating unawareness of - or disbelief in - the fact that the model doesn't record training data vetbatim, but smears it out in high-dimensional space, from which it then samples. The model then doesn't recall past inputs (which are effectively under extreme lossy compression), but samples from that high-dimensional space to produce output. The high-dimensional representation by necessity captures semantic understanding of the training data.

Generating "infinite Doom" is exactly what this model is doing, as it does not capture the larger map layout well enough to stay consistent with it.

Re:Diffusion Models Are Real-Time Game Engines by Workaccount2 on Wednesday August 28, 2024 @10:06AM (#41379689)
Whether or not a judge understands this will probably form the basis of any precedent set about the legality of image models and copyright.
Re:Diffusion Models Are Real-Time Game Engines by znx_0 on Wednesday August 28, 2024 @10:34AM (#41379981)
I like "conditioned brute force" better term.