Is anyone in desktop Linux excited by the possibilities of “AI,” e.g. image generation, large language models, machine learning—especially focused on on-device, responsible implementation?
I feel like it’s been a decade of “well, if we could train an ML model, we could do X…” on the Linux desktop, but now so much of that work is tainted by the current “AI” craze.
Can’t we still have nice “smart” features?
I am not really all that excited yet. The technology needs to advance more before it becomes applicable to the common Linux desktop user.
Most of the AI image generation and LLM options that are available still have heavy resource requirements. It is getting better and there are some options available that aren't as demanding. However the power requirements are still priced out of our reach. Forcing the average user to still have to pay for cloud services in order to accomplish work in a reasonable amount of time.
