brib@bribstodon.xyz ("brib :neofox_floof: :Nonbinary:") wrote:
@firepoet I mean, something I realised about myself is I don't write code for the sake of it... while I like my code to be elegant, it has to be in service of something else otherwise I lose interest too quickly.
I can sympathise with the desire to solve problems rather than write code but I struggle to conceputalise LLMs as a good solution. They produce solutions based on what's in the training data and at that point... wouldn't it be better to improve the underlying software design to prevent unnecessary, code from getting written, rather than LLMs making it brittle?
Having said that I'm sitting on a blog rewrite I haven't tackled yet because it's a huge tasks. Would LLMs make it possible? Maybe, but I don't want to talk to or pay for Clod Clode so the blog's going to sit un-refactored until I find the spoons to rewrite it