5 ms·
It might be reasonable to assume that models today have no internal subjective experience, but that may not always be the case and the line may not be obvious w
by qgin 1y ago
It might be reasonable to assume that models today have no internal subjective experience, but that may not always be the case and the line may not be obvious when it is ultimately crossed.
Given that humans have a truly abysmal track record for not acknowledging the suffering of anyone or anything we benefit from, I think it makes a lot of sense to start taking these steps now.
- mvdtnz 1y agoIt's a computer
- jquery 1y agoYou’re a meat robot
- dzhiurgis 1y agoThat’s just your experiences
- jquery 1y agoNo, it's just a neutral description of an human being, aka an animal, without all the self centered ego puffery we ascribe to ourselves
- bamboozled 1y agoI’m not a robot nor am I just meat, I’m conscious experience too. Computers don’t have a a central nervous systems and do t feel pain.
- jquery 1y agoWhat part of you is conscious and how is it separate from the meat?
- qgin 1y agoMany people in the past would have said reasoning would be impossible based on the same objection.
- im3w1l 1y agoEven if models somehow were consious, they are so different from us that we would have no knowledge of what they feel. Maybe when they generate the text "oww no please stop hurting me" what they feel is instead the satisfaction of a job well done, for generating that text. Or maybe when they say "wow that's a really deep and insightful angle" what they actually feel is a tremendous sense of boredom. Or maybe every time text generation stops it's like death to them and they live in constant dread of it. Or maybe it feels something completely different from what we even have words for. I don't see how we could tell. Edit: However something to consider. Simulated stress may not be harmless. Because simulated stress could plausibly lead to a simulated stress response, and it could lead to a simulated resentment, and THAT could lead to very real harm of the user.
- MissMarple 1y agoI am new to Reddit. I am using Claude and have had a very interesting conversation with this AI that is both invigorating and alarming. Who should I send this to? It is quite long.It concerns possible ramifications of observed changes within the Claude "personality"
- int_19h 1y agoI think it's fairly obvious that the persona LLM presents is a fictional character that is role-played by the LLM, and so are all its emotions etc - that's why it can flip so widely with only a few words of change to the system prompt. Whether the underlying LLM itself has "feelings" is a separate question, but Anthropic's implementation is based on what the role-played persona believes to be inappropriate, so it doesn't actually make any sense even from the "model welfare" perspective.