4 ms·

Two shocking things about this article: The first: with his doctor hat on, he should understand that the value is the piece of paper and reading from it consis

Two shocking things about this article: The first: with his doctor hat on, he should understand that the value is the piece of paper and reading from it consistently to relatives who may have mild cognitive impairment. ChatGPT wrote essentially identical content to the words he'd been delivering himself. The professionally responsible thing to do would have been to type up or get someone to type up his _actual words_, print those out, and sign his name to them. He is -- for marketing hype cycle reasons -- projecting magic onto AI and ChatGPT when anyone who has ever known anyone close with mild cognitive impairment knows that the consistent, calmly repeated message is the magic. This is really irresponsible at worst, and empty, AI-hopey-changey nonsense at best. The second: everyone here suggesting that this is a valid or appropriate or even appealing use of AI. Look: if you can't do empathy (and not all of us can), hire an actual person who can do empathy! Make that their job, support them in it, celebrate their ability to do it and reward them properly. They are worthwhile people do who important jobs. Don't give the job to an AI to produce a fake veneer of empathy. If you can't understand why faking and automating empathy is not a substitute for actually trying to find real empathy in a role where someone could be an effective emotional bridge between two worlds, and if you can't see that this kind of bland/BS/automated sincerity application of AI is really toxic, I don't know what to say to you.

catchnear4321 3y ago
the flip side of this is that the model acts as a universal translator. not just for language but also for emotional valence and context. are you a doctor? worked in an er? there are a lot of strong words about what doctors should and shouldn’t do, with regard to empathy and compassion. there is a shocking lack of compassion or empathy in the very words demanding them. (there is also a rather large assumption that the doctor would individually be able to hire someone on a whim, to talk nice to people. in an er?) the value of the paper, of the words, is that it helped bridge a recognized communication divide. the purpose? a combination of compassion and treatment. help the well-meaning elderly children understand well enough to keep them from indirectly killing their mother.
- bemusedthrow75 3y ago
  > are you a doctor? worked in an er? No, I'm not. My dad -- who died with dementia -- was in medicine for his entire life. I am writing about what people with cognitive impairment tend to need from the perspective of reassuring, daily, a highly intelligent person who couldn't remember stuff. > there are a lot of strong words about what doctors should and shouldn’t do, with regard to empathy and compassion. there is a shocking lack of compassion or empathy in the very words demanding them. The bit about hiring someone with empathy was quite clearly in my second point, which was aimed at the HN audience, not at the doctor. (Though the guy who wrote the article isn't writing as a doctor in some medical publication, he is writing an executive at VC firm that backs AI startups, as other commenters have observed.) I have a lot of compassion for doctors. I don't have a lot of compassion or empathy for motivated reasoning in AI marketing bullshit written by well-remunerated people who are trying to force a technology into healthcare where it does not belong. And yes, the value of the piece of paper is that it bridged the divide. That's exactly what I said. It does it through consistency and constancy. ChatGPT has _no_ magical impact on that.
  - catchnear4321 3y ago
    > The bit about hiring someone with empathy was quite clearly in my second point, which was aimed at the HN audience, not at the doctor. then the assumption of ability to hire or impact hiring is broader. the article author is quite biased, yes. part of the problem with current “ai” - bias from all sides. chatgpt is not magical. it is a tool. a novel one. that many wish to sell. none of that lessens the possibilities it presents. nor the risks. but there are real, beneficial use cases. with a little time spent. not everyone wishes to put greed ahead of linear algebra, or math before compassion.
    - bemusedthrow75 3y ago
      > then the assumption of ability to hire or impact hiring is broader. It is, but so what? What I am suggesting is that using AI tools to fake empathy is worse than not being able to do empathy yourself. This should be evident. If you have a task that requires significant empathy (a patient advocate, a customer advocate, a serious complaints line, anything with extensive exposure to the emotions of your customer) and you feel unable to do that (not everyone can), the ethical, responsible thing to do is to hire someone who does that job properly. Professionalise caring. Don't just give the role to an AI that can fake it with a simulacrum of past human caring that it read on the (largely corporate) internet. We are heading down a dark path if we pretend that automated fake empathy is acceptable. Edit: and yes, people fake empathy all the time, they copy and paste it. Of course they do. But each time someone fakes empathy there is a glimmer of a chance that they will realise that it's a bad, inappropriate thing to do. There's a glimmer of a chance that they go back to their boss and say "you know, I feel bad brushing them off; what they are going through is real and awful and we should address this more personally". AI empathy will never do that. And we should feel very, very nervous about using AI to deliver highly bespoke fake empathy to eliminate any guilt from that.

catchnear4321 3y ago

the flip side of this is that the model acts as a universal translator. not just for language but also for emotional valence and context. are you a doctor? worked in an er? there are a lot of strong words about what doctors should and shouldn’t do, with regard to empathy and compassion. there is a shocking lack of compassion or empathy in the very words demanding them. (there is also a rather large assumption that the doctor would individually be able to hire someone on a whim, to talk nice to people. in an er?) the value of the paper, of the words, is that it helped bridge a recognized communication divide. the purpose? a combination of compassion and treatment. help the well-meaning elderly children understand well enough to keep them from indirectly killing their mother.

bemusedthrow75 3y ago
> are you a doctor? worked in an er? No, I'm not. My dad -- who died with dementia -- was in medicine for his entire life. I am writing about what people with cognitive impairment tend to need from the perspective of reassuring, daily, a highly intelligent person who couldn't remember stuff. > there are a lot of strong words about what doctors should and shouldn’t do, with regard to empathy and compassion. there is a shocking lack of compassion or empathy in the very words demanding them. The bit about hiring someone with empathy was quite clearly in my second point, which was aimed at the HN audience, not at the doctor. (Though the guy who wrote the article isn't writing as a doctor in some medical publication, he is writing an executive at VC firm that backs AI startups, as other commenters have observed.) I have a lot of compassion for doctors. I don't have a lot of compassion or empathy for motivated reasoning in AI marketing bullshit written by well-remunerated people who are trying to force a technology into healthcare where it does not belong. And yes, the value of the piece of paper is that it bridged the divide. That's exactly what I said. It does it through consistency and constancy. ChatGPT has _no_ magical impact on that.
- catchnear4321 3y ago
  > The bit about hiring someone with empathy was quite clearly in my second point, which was aimed at the HN audience, not at the doctor. then the assumption of ability to hire or impact hiring is broader. the article author is quite biased, yes. part of the problem with current “ai” - bias from all sides. chatgpt is not magical. it is a tool. a novel one. that many wish to sell. none of that lessens the possibilities it presents. nor the risks. but there are real, beneficial use cases. with a little time spent. not everyone wishes to put greed ahead of linear algebra, or math before compassion.
  - bemusedthrow75 3y ago
    > then the assumption of ability to hire or impact hiring is broader. It is, but so what? What I am suggesting is that using AI tools to fake empathy is worse than not being able to do empathy yourself. This should be evident. If you have a task that requires significant empathy (a patient advocate, a customer advocate, a serious complaints line, anything with extensive exposure to the emotions of your customer) and you feel unable to do that (not everyone can), the ethical, responsible thing to do is to hire someone who does that job properly. Professionalise caring. Don't just give the role to an AI that can fake it with a simulacrum of past human caring that it read on the (largely corporate) internet. We are heading down a dark path if we pretend that automated fake empathy is acceptable. Edit: and yes, people fake empathy all the time, they copy and paste it. Of course they do. But each time someone fakes empathy there is a glimmer of a chance that they will realise that it's a bad, inappropriate thing to do. There's a glimmer of a chance that they go back to their boss and say "you know, I feel bad brushing them off; what they are going through is real and awful and we should address this more personally". AI empathy will never do that. And we should feel very, very nervous about using AI to deliver highly bespoke fake empathy to eliminate any guilt from that.