I mean they are text apps basically, they should be flying on modern PCs that can do close to a trillion ops per second. Inference is done elsewhere, so what's the deal?
Ask HN: Why are desktop AI apps so heavy?
1–5 of 5 posts
Re: Ask HN: Why are desktop AI apps so heavy?
#2Virtualization missing, that's commonly a difficult task to nail on a messages list.
another thing is mounting and unmounting complex elements efficiently, that was part of why switching threads or sessions felt heavy.
btw, native apps can feel heavy and janky too if not properly optimized for large lists. it's just a matter of shipping a good app that's battle tested with large data sets
Re: Ask HN: Why are desktop AI apps so heavy?
#3Virtualization missing, that's commonly a difficult task to nail on a messages list. another thing is mounting and unmounting complex elements efficiently, that was part of why switching threads or sessions felt heavy. btw, native apps can feel heavy and janky too if not properly optimized for large lists. it's just a matter of shipping a good app that's battle tested with large data sets
[dead]
Re: Ask HN: Why are desktop AI apps so heavy?
#4Static linking, basically: https://en.wikipedia.org/wiki/Static_build
Re: Ask HN: Why are desktop AI apps so heavy?
#5[dead]