• 0 Posts
  • 86 Comments
Joined 1 year ago
cake
Cake day: June 9th, 2023

help-circle
  • The thing is, this pornography and cats will tell future historians a ton about what people were like in our times. Not all of it will be accurate, but that’s an issue with any primary source. Hell, watch some grainy smut from the '70s or '80s and pay attention to things other than the “action”, like the choice of music, the way the actors are talking, how they are dressed, what the sets look like, what kind of excuses for plots are being used, all of which are clearly products of their time. Amateur stuff is even more illuminating. Before anyone thinks I’m overthinking this: We learned a lot about Ancient Rome from the smut Romans carved into buildings in Pompeii.

    It’s the same with old cat pictures. You can reasonably date many of them by what the background looks like, e.g. what kind of electronics and furniture are present, how people who are also in the photos are dressed, image quality (provided it hasn’t been compressed to hell and back since), etc. These kinds of seemingly inconsequential artifacts of our time will be highly illuminating to future historians (provided they are being preserved), just like the complaint letters ol’ Ea Nasir received thousands of years ago.




  • The 1060 is an eight year old mid-range card, lacking almost all of the features that are setting nvidia apart from AMD these days. It has CUDA, but on its own, that’s mostly useful only for non-gaming applications. AMD is lagging behind, but they are not lagging that far behind. AMD has trouble keeping up with 20xx cards and newer, especially when it comes to ray-tracing and upscaling. FSR, while supporting older cards and being manufacturer-agnostic (that’s why even your old Nvidia card is supported), is a crutch that comes with serious visual downgrades, whereas DLSS improves both performance and visuals. This matters in all market segments. Ray-tracing meanwhile is mostly a mid-range and up thing - and while newer AMD cards support it, their performance relative to otherwise equivalent Nvidia cards is lagging far behind.










  • To illustrate your point, my old GPU, a GTX 1080 from 2016 (basically ancient history - Obama was still president back then) remains a very useful for ML-applications today - and this isn’t even their oldest card that is still relevant for AI. This card was never meant for this, but thanks to Nvidia investing into CUDA and CUDA being useful for all sorts of non-gaming applications, the API became a natural first choice when ML tools that run on consumer hardware started to get developed.

    My current GPU, an RTX 2080, is just two years younger and yet it’s so powerful (for everything I throw at it, including ML) that I won’t have to upgrade it for years to come.




  • It could be regulated into oblivion, to the point that any commercial use of it (and even non-commercial publication of AI generated material) becomes a massive legal liability, despite the fact that AI tools like Stable Diffusion can not be taken away. It’s not entirely unlikely that some countries will try to do this in the future, especially places with strong privacy and IP laws as well as equally strong laws protecting workers. Germany and France come to mind, which together could push the EU to come down hard on large AI services in particular. This could make the recently adopted EU AI Act look harmless by comparison.