Soon all that content will be replaced by GPT output, not my doing, but it's happening.
I don't know the consequences of language models recycling their own content, etc. but it's already here. Probably input will be limited to some authenticated trusted sources if possible.
Literally sawing the very own branch they are sitting on.
Hopefully AI can improve ranking systems.
Just guessing.
For example, despite unsubscribing from emails daily and constantly junking, I still get a ton of spam. I assume simply telling a model "never show me a marketing e-mail again" probably works pretty well.
Well, GPT has to learn on something. Right now it's just a language model. It won't get better unless people keep writing these "non-native-english-speaker content".
And maybe we read different things, but I disagree - a lot of my problems were solved by niche non-native-english-speaker bloggers posting about their problems and original research that went into solving them.