5d ago

Is It Just Me?

You're viewing part of a thread.

Show Context

349 comments

Considering the blanket negative post by OP, it's weird to think that OP thinks there's some utility. There's no question of forests or trees, just you putting words in the mouth of the poster.
"Ah yes, AI has many very reasonable use cases and is often quite useful, but is also super evil. I will avoid it and make posts about avoiding it." - you know, a forest of people
What?
- it’s weird to think that OP thinks there’s some utility. [...] you putting words in the mouth of the poster.
  I don't get it. I literally just wrote an entire comment explaining how this was not my claim. Please, point me to where I said that OP thinks there is some utility to AI. Maybe go and re-read my comment in case you missed the whole comment while responding to it.
  
  Let me try again. I also try not to be rude, but a one word "huh?" answer rubbed me the wrong way. Clean slate is fine by me.
  Top Level Comment: "[I found a use for AI.]" My response: "[Although it sounds like you're responding to what OP said, that doesn't really contradict OP since they never claimed it has no uses. They are objecting to AI on different grounds entirely.]"
  I take issue with this bit:
  that doesn't really contradict OP since they never claimed it has no uses.
  Which is the source of your "forest for the trees" comment.
  I don't think that a reasonable person can, based on the reading of the OPs "AI is everywhere and evil" post, presume to say that the OP believed there are good uses for AI. I will probably edit this post for your particular wording. Since in voyager I can't simultaneously reply and look at the comment chain.
  Edit: having read the rant now several times, and in no way is it ever implied in any way that there's anything good about AI. Sure, there's no phrase that you said, or that OP says, that imply that it's explicitly NOT useful. That, however, I would say is the "miss the forest for the trees" part. It strikes me as totally justifiable to say "I have found some uses for AI" in light of a blanket negative OP post. I don't think the guy you initially replied to is missing anything.
  Hopefully that fully articulates my position.
  
  Ok, I'll admit, "huh?" was terse and I should have been more polite. In my defense, I thought you were not arguing in good faith given that it felt like you were putting words in my mouth -- you said I claimed the author thought ¬X, but what I said was "the author did not say X." But yeah, I was rude -- I'm sorry.
  Okay, I understand your position better, but I still disagree with you. I don't claim to know whether the OP thinks there are or are not useful uses for AI; rather -- and this is where we apparently disagree -- I don't think that "there are some good uses for AI" is a cogent rebuttal to the OP's claim. This is because I don't think it makes sense to look at a list of things that OP says, then imagine an additional point that the OP would likely agree with based on their overall vibes (but is actually entirely logically independent of everything they did say), and then refute that point.
  For what it's worth, I am a person who agrees with most of what the OP says but I still wouldn't claim that AI is entirely devoid of utility. If I were someone who used microblog social media, I could easily see myself posting a similar rant as OP, since I have strong negative feelings about AI, but I would be mildly peeved if somebody responded the way top-level-comment did, since it's not a response to anything I said, nor even a point of contention for me. (I can enumerate my grievances with AI if you wish, but I don't think it's really important.)
  
  I also apologize.
  I guess we have to agree to disagree and agree to agree? LLM design happens to be my area of work for the last 4 months or so and I tend to agree that it's a hateful domain without much to offer. That said, my ability to do my job is extremely enabled by perplexity.ai. I think about 95% of regular people's queries to generative AI of any kind is a waste of effort an active pursuit of brain atrophy, but there is definitely some upside.
  I don't think that sentiment is echoed by the original post. I also think that with the overwhelming number of anti/fuck ai posts it's worth highlighting upside when there is upside. Last, I just don't in any way think that someone who writes a full bore "ai is evil" post has left any space to discuss any sort of benefit. So that's my main disagreement with you. I think "AI is stupid" is implied if there's no single positive mention.
  I can honestly say perplexity.ai has saved me hours of technical research debugging, and I've got a PhD in rocket science and Ive been programming since 2003. By and large it's totally stupid, but if you know what you need and what good answers look and feel like, it's hard to be like "AI BAD!" which is the majority perspective of this type of thread.
  
  Mkay, I think we agree more than we disagree. For instance yeah, I agree that AI has its uses and I actually agree with the top-level comment. Like you, I also don't particularly wish to join in the chorus of un-nuanced "AI bad!" takes (as opposed to more nuanced anti-AI takes). If there's a post that's exaggerating the environmental impact of AI, I will upvote a well-researched comment that dissents. (Environmental impact strikes me as one of the places where AI fears are really overblown -- golf courses currently use over 15× more water than datacenters in total, but they don't have even a 15^th the outrage.) So yeah, I would rather that the "fuck AI" crowd be more level-headed when agitating against AI and not use poorly-researched, poorly-reasoned arguments when there are much better arguments against AI. If you were to really boil it down, I would say I agree with you here:
  I also think that with the overwhelming number of anti/fuck ai posts it’s worth highlighting upside when there is upside.
  But it's the same principles that lead me to upvote thoughtful dissent that makes me dislike the top-level comment here. It's not thoughtful. It might be a good comment in a thread about how AI is completely worthless with no redeeming qualities -- and I'd upvote it there -- but I am not going to take my frustration with threads like that and upvote the comment in this context, where a dozen other reasons were given to rally against AI (some I agree with, some I don't) and top-level-comment chooses the one time to bring up a good point in a context where it's not relevant.
  Sure, the OP here is being rather overzealous, and if I were going to leave some dissent in this thread I would choose one of the things they said I disagreed with. But I just don't think this thread specifically is an appropriate place to say "actually, AI has some good uses." That strikes me as rounding the OP toward zero, and we don't need more of that.

349 comments