A new subject for a PhD 
What is your point?
That studies would need ro be conducted and peer reviewed, otherwise it is just assumptions.
And on top the question is if LLMs are the next communicative evolution that would majorly impact humanity anyway.
There have always been major changes if new types of media have been introduced.
I think we all know LLMs will certainly stay.
Remember when internet addiction was a thing? Boy, have things changed!
The underlying psychopathy hasn’t changed … just the symptoms 
In my estimation AI will never be a thing because of it’s tendency to hallucinate and tell whoever trains it what they want to hear
" Yeah … meet the new boss … same as the old boss" — The Who 1971 wisdom
I think it’s easy to underestimate just how deeply and rapidly AI is already becoming embedded in our world. While hallucinations and biases are real challenges, they’re being actively addressed, and in many areas AI is already proving useful (medicine, logistics, engineering, design, customer service, even scientific research). The trajectory suggests it will be woven into daily life in ways most people can’t yet fully comprehend.
These are factual areas that AI can access real factual data … I think what the original topic addresses is more “ephemeral” human subjects that AI will never understand because it is incapable … it simply regurgitates information and cannot come up with anything original … only what it is told is “true”
And who is addressing these exactly? … the same ones who mine our data, extract every bit of coin of our pocket and say “just trust me bro”
Yes, my comment was not about using for AI social interaction. The biggest question of our time is “What is it to be human?”. Human to human interaction is a vital for our own well being IMHO. (The idea of AI driven robots is scary to me.)
And that is what the internet and “social media” has killed … we need to get back to community things … Thankfully I live in rural area where we still do such things
I always am very skeptical and cynical from reading this kind of statements. Are we talking about the small rural communities where LHGTB+ people are bullied out, where marrying a protestant instantly makes you lose all your catholic custmomers, where refugees are attacked, where gossip is the main source of information, where racism is an accepted thing, where even thinking differently about whatever leads to exclusion?
Oh, yeah. I do agree to that. That said, from experience I can tell that video conferencing can come pretty close to person to person, since you hear a voice, tone, you see facial experessions, body language to a certain extent.
Well about the same kind of scary as so-called religious leaders IMHO.
I use and find AI bots a bit useful for technical output. Together with a very critical view of the output, with massive testing of the output, and multiple iterations, I find such works well.
I find I can use an AI bot and it can be useful to dig out obscure information on an old news story, that when searching I can NOT find with google. However the AI Bot has a tendency (IMHO) to fabricate/make up false answers. Everything it produces needs verification. More than once I have taken the specifics of an AI bot answer and fed it into Google and did a search (rephrased such that a Google search possible - but such Google refined search only possible due to AI bot ‘hints’ ) and confirmed or disproved the AI bot answer.
And often I will take an AI bot’s (old news story - or research output) feed it into other AI bots and watch them tear the first AI bot’s statement apart - or simply say there is no such thing/news.
Still it can be useful - albeit massively healthy skepticism is needed.
But to use an AI bot for more than the above (more being such as using an AI bot in social or moral aspects or philosophical aspects) I currently believe a total waste of time. … Perhaps I will be proven wrong eventually, but at present time I find a couple of ‘niche’ areas where an AI bot can be useful , and for all other aspects I am to a large extent “from Missouri”.
Talking to a bunch of scientists (politicians, gurus, …), studying a thousand books in a library, searching the internet or asking ChatGPT one must always keep in mind:
- Humans do have strengths and weeknesses,
- Every human being makes errors.
- All humans do have their individual agenda.
- Everything a human being does is in the end driven by the aim to bring its own agenda into live.
So every tool (e.g. AI, internet search engines, books, newspapers, …) created and every answer provided by human beings are shaped by those human features. The real problem is that today many people just accept information as “true” or “correct” without considering this.