I think we are talking of the same thing, but from different perspectives.
If I understand you, you are talking of non-AI-generated social media content posted by and interacted by humans vs the AI-generated junk created by AI bots & script kiddies. You probably feel the internet is getting diluted as the authentic (human-created) content is replaced or superceded by junk stuff (AI generated or trolling content) that are more irrelevant fluff & unavoidable diversions than the real raw human-driven content and exchange of ideas spurring meaningful dialogues.
I am talking of same human created social media (and news media, and info-repository media (including wikipedia, books, journals, etc.)) content created and maintained by humans, which unfortunately tend to have inherent bias & skewed ideologies leading to bad/unwarranted behavior.
But when all that online and offline content is consumed as data repository and training material for AI LLMs, then the AI bots will inherit and demonstrate those same biases, skewed ideologies and resulting bad behavior.
e.g., the main subreddit (under Reddit.com) for my nation, is operated by admins belonging to enemy nation. They use bots for such propaganda. Reddit knows this problem, but hasn't bothered to do anything about it, because that bias and malice against my country and its culture is widespread among Leftists, who own and control most of social media, news media, info-repositories (like Wikipedia, which too openly spreads such propaganda). So that subreddit is filled with hate posts and malicious misinformation about my country and culture.
Last year, Google made a deal to absorb all Reddit content, in order to train it's AI LLM models. Now guess what those AI LLMs will think and say about my country and culture?
You miss the real human content and realistic online interactions, but I feel you are missing the fact that such content will always have bias and driven by ideologies or opinions. But perhaps that difference in ideas and opinions is what you seek. You are probably afraid that AI will replace such content creators and content, thus leading to boring online communities and lack of meaningful interactions.
I am afraid that such human created content and bias will inevitably percolate into AI LLMs content and behavior, and it will forever skew the world's opinion about my nation and its ancient culture.
Unfortunately , it looks like AI will reshape our online realities and interactions to make it all a worser place from which there's no return or escape (AI will become all-pervasive in our lives, whether we like it or not).