Reduce labor costs to increase profits. But that's for the good of a small number of people when all meaningful work can be automated. And I don't trust the billionaires to make society better for the rest of us.
Even worse: Who will produce new training data? If everyone only uses AI, then nothing new will be invented. Quality will stagnate or go downhill. It cannot get better
But haven't you heard? AI gets better every month! It's just like Moore's Law and it will keep getting better because why wouldn't it? It's not like there is a dead-end to something fundamentally built around categorization and regurgitation of existing things.
(Noting this is all sarcasm in case I didn't lay it on thick enough)
To me "learn to code" has ALWAYS been a synonym to "learn to think". Coding is nothing else than thinking. You learn a handful of structures and then you only think about how to combine them. It doesn't matter if you do it by AI or "by hand".
Sounds interesting, but since I haven't been in touch much with this topic I ask myself: Does this have any benefit for my personal home-computer usage?
For a long time I have the urge to try out Nix, because I clutter up my computer way too fast and therefore often get mad and just install a fresh system. This works fine with my files, but there are always applications which I forget about and forget to save configs. So having this all in a git repo to spin it up fast would be nice. Is bootc, fedora silverblue and so forth trying to achieve something similar?
I think bootc is exactly what you're looking for. I use it[1] for configuration like you mentioned but also for:
- Installing codecs from third-party repositories. This is especially nice to do in CI because you get a build failure if packaging drift happens.
- Installing out-of-tree drivers. Again, you get a build failure in CI if an out-of-tree kernel module won't build. In addition, you can use multi-stage builds (see the Dockerfile in my repo for an example) to avoid pulling dependencies into your final system image. This saves me from having the 70 or so RPM packages that are required for building NVIDIA drivers installed on my PC.
It's not as ambitious as NixOS but I think it gives a lot of the same benefits with far less effort.
Still waiting for ROCm on my cheap Radeon RX 7600. Would be nice to play around with it a little. I know that this card is nothing fancy. There is somewhere a github issue where they announced to port it for linux to consumer cards, but last time I checked (a few days ago) it still wasn't available
You should be able to make it think you have another card:
export HSA_OVERRIDE_GFX_VERSION=10.3.0
The possible values are said to be:
# gfx1030 = "10.3.0"
# gfx900 = "9.0.0"
# gfx906 = "9.0.6"
# gfx908 = "9.0.8"
# gfx90a = "9.0.a"
Telling ROCm to pretend that your RDNA 3 GPU (gfx1102) is an RDNA 2 GPU (gfx1030) is not going to work. The ISAs are not backwards-compatible like that. You might get away with pretending your gfx1102 GPU is a gfx1100 GPU, but even that depends on the code that you're loading not using any gfx1100-specific features. I would generally recommend against using this override at all for RDNA 3 as those ISAs are all slightly different.
In any case, the possible values can be found in the LLVM documentation [1]. I would recommend looking closely at the notes for the generic ISAs, as they highlight the differences between the ISAs (which is important when you're loading code built for one ISA onto a GPU that implements a different ISA).
Use the PyTorch Nightly build. The ROCm libraries themselves have been built for the RX 7600 (gfx1102) since ROCm 5.4/5.5, but PyTorch itself wasn't enabled until a few weeks ago. The RX 7600 is still not 'officially supported' on Linux, but I have an RX 7600 XT and I haven't encountered any issues in my (admittedly intermittent) use of the card in AI applications. You may, however, find the 8GB of VRAM in the non-XT version to be a limitation.
