@thomasfuchs@hachyderm.io
Itβs always βAI is great for generating boilerplate codeβ and never βwhy do we even need boilerplate code, maybe programming is brokenβ
@david_chisnall@infosec.exchange
@thomasfuchs@hachyderm.io
The one good use Iβve seen for LLMs in programming is to spot missing abstractions. If an LLM can generate code for a particular task, it means that this task has been repeated enough times that it should be factored out.