Follow
@Lioh I kind of see a little irony in your post, like: solving problems that shouldn't exist...
Though I am trying to get my foot into llm as well. With ollama its not the same experience, but at least I have full control when it comes to running the model, both in terms of power consumption and privacy. Think that should be the way to go.
@ovid @Lioh I was specifically refering to liohs communication example. If we have to use AI to communicate to the gov, there is still something wrong.
But other usecases as yours are absolutely valid imho.
I just think we should't use glue because its easier than screws if we still want to disassemble things.
@petrisch @Lioh Solving problems that shouldn't exist? I see this crop up from time to time, but when I use AI, I'm solving plenty of real problems that *do* exist and won't go away.
For example, one of the more common cases is when non-native speakers use it to write email that sounds like it's by a native English. It's faster than writing it manually, has fewer errors, and makes communication easier. Also stops many prejudice issues.