>>106211723
ChatGPT makes legal decisions, perpetually interpreting local laws and providing or withholding information according to context. Classifying legality of text is just one task comprising the massive range this technology is capable of that would decrease labor in specifically the legal field. Parsing tens of thousands of documents simultaneously to derive context for a case reduces labor by at least a third. In terms of medicine, neural networks have already designed dozens of drugs. Massive companies were being established around this implementation of neural networks years before LLMs.
>Insilico
>Exscientia
>Atomwise
AI can classify fractures, tumors, trauma, ruptures- all at rates more accurate than human doctors. They will likely use these tools to classify such, lowering the labor requirements.
>>106212398
>construction
There are several implementations of drone bricklayers, or even entire houses 3D printed with concrete. Besides that, putting a nail in wood is left to humans but the same can't be said for
>welding
>excavation
>anything sufficiently dangerous, raising costs
>quality assurance
>>106215807
AI itself approaches the point of perfecting AI. For example, generative weight initialization, generative model structure through problem schematic analysis, and even the now brute force design of entire models and pipelines by LLMs.
>Learn AI architecture
I'm completely inclined to agree though, because I know this technology will completely wash away all familiarity of today's society, which I existentially despise.
