After working at a considerable number of companies it's clear that most software work is to a large degree just replicating things that have been done other pl

After working at a considerable number of companies it's clear that most software work is to a large degree just replicating things that have been done other places, and not like solving the same complicated interesting problems, but figuring out how to do their own local variation of plumbing various things together. Many businesses also only have a competitive advantage because of some set of hidden knowledge. And not "we spent a lot of work inventing this new thing" but "we have contacts and knowledge of this poorly documented messy system and we only exist because of it's continued state of brokenness". The real threat of AI/ML/LLMs are that they're going to get smart enough to do all of this stupid incredibly repetitive work and then there's not going to be much left to do. We might actually reach a post scarcity economy and the transition will be _interesting_.