Hacker News Reader

Comment by LoganDark | Hacker News Reader

No, the benefits of AI do not mean anything like we have to accept AI slop contributions to open-source projects. We can still shun the blatant abuse. We don't have to just shut up about it because we might stop frontier development. We won't.

To me, "accept" means to socially tolerate it. No. Bad. It's not socially tolerable. It is still wrong to use LLMs for evil. I don't think that means LLMs are a fundamentally evil technology, but that does NOT mean anyone should get a free pass to use them FOR evil.

View on HN

https://yourlogicalfallacyis.com/strawman

They're twisting my definition of the word "accept". I was clarifying what I meant with an example. I think instances of evil behavior need to be shunned regardless of the other benefits of the technologies being used for evil. No one should get a free pass to do evil based on the technology.

I also believe the technology itself deserves to be improved so that it can be used for good. That doesn't mean I want more evil or to be more accepting towards evil. Evil should still be shunned.

Can you expand on this? It is not clear what part of the post you regard as attacking a strawman.

It probably looked like I was equating "accepting the tradeoffs" to "condoning LLM slop contributions to open source projects". I was referring to enabling individual instances of abuse though, not accepting the technology overall in spite of the bad behavior. It's enabling the individual instances that I'm protesting, because I firmly believe we shouldn't. I should've been clearer, my bad.

The parent comment was:

> > This is kind of how the world works. We have a set of tradeoffs to make and the downside of the tradeoff we accept in exchange of the advantages.

The reply was:

> No, the benefits of AI do not mean anything like we have to accept AI slop contributions to open-source projects.

The parent at no point said the we have to accept AI slop.

Reply on news.ycombinator.com