-3
AI Chatbots, if created a Non-Profit Organization, would actually be a useful tool for loneliness/depression without judgement. (piefed.ca)
submitted 1 day ago by to c/unpopularopinion@lemmy.world
14 comments fedilink hide all child comments
 

It it not actually unpopular in the broader population, but AI seems like a dirty word on lemmy, just like a far-right's gut reaction to words like "communism" or "socialism".

all 14 comments

sorted by: hot top controversial new old
[–] 14 points 1 day ago (8 children)

Ah yes, those chatbots that helped people plan their suicide. Very helpful and in no way a flawed technology

  • permalink
  • fedilink
  • source
  • hideshow 8 child comments
    •
  • [+] [S] -8 points 1 day ago (7 children)

    Yes, thats why corporate control is bad. The problem is corporations, not the tech itself.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 7 child comments
    •
  • [–] 10 points 1 day ago*

    There is no way to control chatbots enough to make them safe, even if they weren't under capitalist control the technology just does not work that way. An emotionally vulnerable person, one who's lonely or depressed, is exactly who you don't want relying on a chatbot. Whether it spins into some ego trip or encourages suicide or steers them towards murder. Whatever it is there's no thought there, it doesn't understand it's words. There's zero way to keep it in check. The "guardrails" marketing materials claim are on chatbots now are just suggestions to the things.

    That's why the "my grandma" hack and such worked so well, because the thing has zero understanding. It doesn't know it's causing harm. Even one created and controlled by people who genuinely mean well it would not be a good idea.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 6 points 1 day ago (5 children)

    There is no substitute for professional help. AI for depression and lo eliness is like heroin for pain management - feels good, but isn't doing you any favours and might just be the thing that ends you.

    A Psychiatrist isn't judging you, you are getting a diagnosis, therapies and treatment. It doean't always feel as easy, because help isn't always easy or comfortable.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] [S] -1 points 1 day ago* (last edited 1 day ago) (4 children)

    Unfortunately "professional help" is not some deity that's above human bias.

    For the medicinal aspect, all they do is keep trying a bunch of med that do not work.

    For the talking aspect, its nothing very useful, regardless of who it is. Is the "oh yeah that must be so hard for you", either read from an instruction manual or reduce things down to "just go outside and take a walk" "just do x". Meanwhile the front desk person tells you "hold on, you got a bunch of copays, go to the billing office" WHAT, when was that a thing? They never even said anything for the past several times. Pretty sure the insurance was supposed to cover that.

    And frankly, your comment seems very Eurocentric. Like EVEN in a western democracy, healthcare (ahem 'Murica) isnt free. Disability Medicaid/Medicare (whatever the term is) is not easy to get. Its a bureaucratic nightmare.

    And think about all the countries in um...

    Africa

    Middle East

    Southeast Asia & South Asia

    LOTS of places dont have the luxury of healthcare, and they have cultures that actively stigmitize mental health. Like I know my parents homeland, People's Republic of China, has a horrible stigma of mental health. An open source AI model would be REALLY HELPFUL for those in these ultraconservative societies.

    (Of course beta software in testing isnt stable, but if we have EFF or Wikimedia Foundation develop it (example organizations, not saying they would be the best to develop such a thing), eventually we could have a stable usable model)

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 2 points 1 day ago (2 children)

    For the talking aspect, its nothing very useful, regardless of who it is. Is the "oh yeah that must be so hard for you", either read from an instruction manual or reduce things down to "just go outside and take a walk" "just do x".

    It sounds like you have no idea WTF you're talking about. This is the response I would expect from someone that has just heard from other people that have never gone to a therapist, what therapy is.

    I'm certainly not a proponent of AI, but AI for therapy is arguably the absolute worst use case. Therapists and psychiatrists are medically trained to work in their professions. As far as drugs, yea many times it's not the right drug it disagree on the first attempt because everyone is different and not everyone responds the same way to the same drug or dosage.

    Depression is a condition that is crippling, and for some, anti depressants are the solution due to chemical imbalances. I personally know a handful of people that have gone through the ringer to find what works for them, and they are still alive today because of it. No AI can do that.

    I do agree that the cost is an unfortunate hurdle, and I personally have been affected by that. Some non empathic and non human AI that will just add likely tell a person to kill themselves as give God advice is not a solution.

    If you had cancer, would you turn to some non profit AI to give you a treatment, or would you go to an oncologist instead? It's the exact same thing when dealing with mental issues.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] [S] 1 point 1 day ago (1 child)

    I dont argue that AI is a cure, its more like a bridge to feel more stable and so they can be mentally sane enough to continue to seek self improvement, whether it be going outside, trying to socialize at a club, or employment. Because having an AI, feels like having "someone being with you" so you wouldnt feel anxious at the world. At that would be least if such an AI could be perfected, which of course currently these corporate backed AIs aren't nearly as good as it could be, on top of the privacy issues.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 2 points 1 day ago

    feels like having "someone being with you"

    This is the problem. AI isn't that. There is no knowledge in an LLM. It's impossible to train them in such a manner as how you see them being beneficial. Using it as a bridge is like getting blackout drunk to get to the next day. For the people that think they're finding solutions, it's more of an addiction.

    its more like a bridge to feel more stable and so they can be mentally sane enough to continue to seek self improvement, whether it be going outside, trying to socialize at a club, or employment.

    This is therapy in a nutshell. Being able to do that effectively takes special training and situational awareness that LLMs can't do. In your last post, specifically the part I quoted in my response, was you griefing therapy for exactly what your saying LLMs can do.

    I encourage you to look into the real world anecdotes of people that have used AI for this purpose, or the people that have created significant others.

    It would be great if it were possible for someone that's struggling to ask a question of their computer and get all of the answers, but that's nothing but wishful thinking. Even if the perfect model were trained with the perfect material, and somehow had no bias, it still wouldn't be a bad choice, inherently because the technology just isn't capable of doing what you're looking for. It requires a human mind with empathy and true understanding to be able to guide someone, not a computer that is finding the closest word match based on a users input. There's no intelligence or understanding in AI, just best guess to the next word.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 1 point 1 day ago

    I see where you are coming from and you're not wrong, well not exactly. You have to understand that Psychiatrists not being infallible is correct, but pointless. They are the best, most reliable help. Say what you want about the limits of modern medicine, it is by far the best thing ever invented and keeps getting better.

    The meds take time and its a trial and error process. Patience is a virtue. While you figure out what works best for you, some help, some hurt, and its up to you to find which is which. What we know is that it is not just a deficiency in neurotransmitters. A pill can't fix everything. But its part of a possible solution.

    Going for a walk is not dismissing things and you should not dismiss it. It is backed by hard science and costs nothing. Getting exercise is a proven reliable thing. Do it, regularly. It is not a cure, but it does disrupt the doom looping rumination that characterizes depression. Treat it like a prescription, powerful medicine with no copays. Go forest bathe. Practice meditation like nature appreciation. Neuroplaticity is an ability to rewire your brain.

    Those three pillars in the right amounts (meds, exercise, cognitive reinforcement) are proven to help manage depression.

  • permalink
  • fedilink
  • source
  • parent
    • [–] 3 points 1 day ago

    Why do you care what "Lemmy" thinks?

  • permalink
  • fedilink
  • source
    •
    [–] 2 points 1 day ago (1 child)

    I don't see how the non-profit is relevant there. If it works then it works independent of who made it. I just don't think we're there yet. If you're stranded on a deserted island then maybe it'll do a better job keeping you sane than Wilson does but in its current form it's nowhere near the replacement for real human connection.

    The suicide examples people love to bring up are not because the evil corporations made them such - it's an inherent flaw of the systems themselves.

  • permalink
  • fedilink
  • source
  • hideshow 1 child comment
    •
  • [–] 1 point 1 day ago* (last edited 1 day ago)

    Yes. We all train chatbots the same way. There's a first step where it's fed a lot of text to learn to mimic coherent language and learn facts. And then there's a second step where we tune it to do the thing we want it to do. For example write dialogue, answer questions... In contrast to for example write a novel or write unproductive Reddit arguments.

    All of this happens disregarding of who does it. A commercial business or a non-profit. And you'd be left with a similar result with all the upsides and downsides of a chatbot. It will hallucinate, make people spiral into stuff, give bad advice... Amplify mental disorders. That's just what chatbots do, due to how they're made. And there's currently no way around it.

    So it really doesn't matter all that much who does it. It'd help if we had a different type of chatbot which isn't subject to the same kinds of limitations. But we don't.

  • permalink
  • fedilink
  • source
  • parent
    • [–] 2 points 1 day ago

    I don't agree with the typical attitude of unconditional rejection of AI, and find it pretty useful overall. But I am skeptical this is a good thing for it to be used for, because a big part of the problem and cause of loneliness and depression is all the ways our society is structurally isolating people and cutting them off from each other's support, which is itself a larger problem than the mental distress it produces. Ways of partially alleviating that distress which cannot result in any sort of human connection (not just AI companions, there are lots of things that do this to varying degrees) I think are ultimately making it worse, because support isn't one-way, if someone is cut off from others, others are also cut off from them.

    Your computer is still just your computer, it isn't capable of being your friend, because it isn't a person, even though it's better than a corporation, which isn't a person and also has bad intentions towards people. The best it can do is help you communicate with others.

  • permalink
  • fedilink
  • source
    •
    [–] 1 point 1 day ago

    I mean if the chatbots made all the lonely and depressed folk off themselves, that techinically would fix the issue for like 5 minutes, until the new folk get depressed and lonely.

  • permalink
  • fedilink
  • source
    •