What's scary is how many people seem to actually want this.
What happens when hundreds of millions of people have an AI that affirms most of what they say?
What happens when hundreds of millions of people have an AI that affirms most of what they say?
Lots of them practiced - indeed an entire industry is dedicated toward promoting and validating - making daily affirmations on their own, long before LLMs showed up to give them the appearance of having won over the enthusiastic support of a "smart" friend.
I am increasingly dismayed by the way arguments are conducted even among people in non-social media social spaces, where A will prompt their favorite LLM to support their View and show it to B who responds by prompting their own LLM to clap back at them - optionally in the style of e.g. Shakespeare (there's even an ad out that directly encourages this - it helps deflect alattention from the underlying cringe and pettyness being sold) or DJT or Gandhi etc.
Our future is going to be a depressing memescape in which AI sock puppetry is completely normalized and openly starting one's own personal cult is mandatory for anyone seeking cultural or political influence. It will start with celebrities who will do this instead of the traditional pivot toward religion, once it is clear that one's youth and sex appeal are no longer monetizable.
Social media follows a similar pattern but now with primal social and emotional circuits. It too causes troubles, but IMO even larger and more damaging than food.
I think this part of AI is going to be another iteration of this: taking a human drive, distilling it into its core and selling it.