3 ms·

> “Interestingly I felt heard & warm. Never tried therapy before but this is probably it? Try it especially if you usually just use it as a productivity tool.”

> “Interestingly I felt heard & warm. Never tried therapy before but this is probably it? Try it especially if you usually just use it as a productivity tool.” If you’ve never tried therapy before, how can you possibly know if AI is replicating it. Maybe try actual therapy before recommending AI therapy to others? I’ve personally had some pretty uncomfortable therapy sessions at times, because I had some pretty uncomfortable life experiences - of which ChatGPT wouldn’t even be allowed to discuss with me because the content of my experiences would breach it’s content policy. So maybe it’s good for therapy if all you need to do is talk about how busy it is to have to work, and also do other adulting things, and you just need someone to indulge your self-pity. I’m super skeptical about this. > “Peer-to-peer mental health app Koko decided to experiment with an artificial intelligence posing as a counselor for 4,000 of the platform’s users.” Is that even ethical? > “On the darker side of things, earlier this year, a Belgian man’s widow said her husband died by suicide after he became engrossed in conversations with an AI that encouraged him to kill himself.” Sounds like something Replika would do.

mindcrime 3y ago
“Peer-to-peer mental health app Koko decided to experiment with an artificial intelligence posing as a counselor for 4,000 of the platform’s users.” What I find discouraging is this: A while back (even before the current LLM / GPT craze started) the idea briefly flickered through my mind of having an "AI therapist" as a chat-bot. I thought the idea was so obviously bad in so many ways that nobody would ever take it seriously. To sanity check myself, I started telling a friend about the idea, and before I even finished talking, they were like "absolutely no. No way. Horrible, horrible idea. Never, ever do that." Which is what I expected and hoped to hear. I pretty much forgot about the whole idea, because - again - it's such an obviously bad idea. No, not just a bad idea, it's a horrible idea. And now, here we are. Somebody (or multiple somebody's) have actually started down that path. These are the times when I want to hang my head in shame and pretend not to be affiliated with the tech industry in any way. :-( Is that even ethical? I'm gonna say "fuck no."
- salawat 3y ago
  I have had the chagrin of watching almost every example of bad ethics we delved into in my "Ethics of Computer Science" coursework become the basis of multi trillion dollar companies within the last decade or so. Might as well have called that course "Forecast of the next decade and a half's highest paying jobs". In fact, if anything, this has led to a massive loss in faith on my part in our system being in any way incetivized to select for ethical behavior.
- Quinzel 3y ago
  I do wonder what the arguments for AI therapy are Vs the arguments against AI therapy. I am in agreement with you, I just intuitively think it’s a terrible idea - however, my intuition could be wrong. I think I’d be more comfortable making an argument against AI therapy, but I think it would be important to hear arguments for AI therapy as well. Is it possible that AI could be used to enhance therapy, rather than replace it?
  - mindcrime 3y ago
    To me, probably the biggest argument against is is the privacy/confidentiality aspect. I mean, OK, sure, conversations with a real world human therapist could be leaked. I'm sure it has happened... but I believe the odds of your AI-Therapist's server being hacked and having all of the chat logs leaked is a lot higher than somebody breaking into your IRL therapist's office and stealing their notes, or water-boarding them into giving up all the salacious details about your life. Outside of that, I'm not sure I believe that AI's will be able to delivery truly efficacious (and safe) therapy anytime soon. And don't get me wrong - I'm an AI advocate. I spend a lot of my time on AI and related subjects, and believe AGI is possible, maybe even probable. But even AGI won't necessarily understand human feelings and emotions since at the end of the day, the "A" in "AGI" still stand for "Artificial". That is to say, I think we can make machines that are "as smart as humans" without being human, or having a shared, visceral understanding of the human experience. But mostly, I expect that "AI therapist" services will just be data logging / data mining fronts that will use the content of your most private thoughts to try to sell you more shit, and/or will eventually get hacked and leak your most private thoughts onto the 'net. So yeah... not something that needs to exist, IMO.

mindcrime 3y ago

“Peer-to-peer mental health app Koko decided to experiment with an artificial intelligence posing as a counselor for 4,000 of the platform’s users.” What I find discouraging is this: A while back (even before the current LLM / GPT craze started) the idea briefly flickered through my mind of having an "AI therapist" as a chat-bot. I thought the idea was so obviously bad in so many ways that nobody would ever take it seriously. To sanity check myself, I started telling a friend about the idea, and before I even finished talking, they were like "absolutely no. No way. Horrible, horrible idea. Never, ever do that." Which is what I expected and hoped to hear. I pretty much forgot about the whole idea, because - again - it's such an obviously bad idea. No, not just a bad idea, it's a horrible idea. And now, here we are. Somebody (or multiple somebody's) have actually started down that path. These are the times when I want to hang my head in shame and pretend not to be affiliated with the tech industry in any way. :-( Is that even ethical? I'm gonna say "fuck no."

salawat 3y ago
I have had the chagrin of watching almost every example of bad ethics we delved into in my "Ethics of Computer Science" coursework become the basis of multi trillion dollar companies within the last decade or so. Might as well have called that course "Forecast of the next decade and a half's highest paying jobs". In fact, if anything, this has led to a massive loss in faith on my part in our system being in any way incetivized to select for ethical behavior.
Quinzel 3y ago
I do wonder what the arguments for AI therapy are Vs the arguments against AI therapy. I am in agreement with you, I just intuitively think it’s a terrible idea - however, my intuition could be wrong. I think I’d be more comfortable making an argument against AI therapy, but I think it would be important to hear arguments for AI therapy as well. Is it possible that AI could be used to enhance therapy, rather than replace it?
- mindcrime 3y ago
  To me, probably the biggest argument against is is the privacy/confidentiality aspect. I mean, OK, sure, conversations with a real world human therapist could be leaked. I'm sure it has happened... but I believe the odds of your AI-Therapist's server being hacked and having all of the chat logs leaked is a lot higher than somebody breaking into your IRL therapist's office and stealing their notes, or water-boarding them into giving up all the salacious details about your life. Outside of that, I'm not sure I believe that AI's will be able to delivery truly efficacious (and safe) therapy anytime soon. And don't get me wrong - I'm an AI advocate. I spend a lot of my time on AI and related subjects, and believe AGI is possible, maybe even probable. But even AGI won't necessarily understand human feelings and emotions since at the end of the day, the "A" in "AGI" still stand for "Artificial". That is to say, I think we can make machines that are "as smart as humans" without being human, or having a shared, visceral understanding of the human experience. But mostly, I expect that "AI therapist" services will just be data logging / data mining fronts that will use the content of your most private thoughts to try to sell you more shit, and/or will eventually get hacked and leak your most private thoughts onto the 'net. So yeah... not something that needs to exist, IMO.