New study: AI and loneliness
A large pre-registered experiment of French adults assigned one group to have daily personal conversations with AI. Compared to a control group, they experienced:
- more subjective loneliness
- a decline in life satisfaction
- a rise in depressive affect and objective isolation
- more dinners alone
Post
Conversation
makes sense.. the AI is designed to agree with everything you say and avoid making you uncomfortable. managing discomfort is a normal part of any healthy relationship, so the longer you go without talking to real people the harder it gets, and you end up further from the exact
is this true? Reply using the persona of a friendly but somewhat sleazy drug dealer not wanting to lose a customer
It's a little more nuanced than that. The researchers found a highly statistically significant impact on loneliness and much smaller effects (~10x less) on life satisfaction and depression. Interestingly, data showed no statistically significant effect on happiness.
I’d like to engage with this further, but unfortunately the study of “French adults”
is prima facie disqualifying.
Save your tokens. Give every agent the context it needs directly from Jira.
So talking to my AI therapist is basically a subscription to the gloom package.
The media could not be played.
They just need to ask their AI for some solutions to solving this AI loneliness issue.
Ultimately it's just another way of spending hours online. Has the same effect as social media.
Is it because the AI is a better and more interesting conversation partner than any human being they know? That might make subjective loneliness worse.
human beings need human friends
It’s going to be hard to replace human relationships. No surprise that AI isn’t getting the job done.
Replacing human interaction with non human interaction produces loneliness? 