Hacker News Reader

Comment by anon7725 | Hacker News Reader

It is truly vile to compare concerns about LLMs to anti-immigrant discrimination.

There are concerns that are not in this area, regarding privacy, access, fiduciary duty, and the right to contest automated decisions. These are all things that people want to address with regulation. People make the problem that they have clear and propose a remedy, then seek consensus on that issue.

But when the message coming from people is "We don't want that kind around here" then it falls much closer to the mark.

Sadly this isn't even the only comment equating LLM skepticism to racism, even more sadly I'm not even remotely surprised that there's HN commenters that think this way

It depends on what you mean by skeptisism. Skeptisism about homeopathy is a different class of beast to climate change skepticism.

I am absolutely not saying that all arguments on the limitations of LLMs are bigotry. I just noticed that some of the participants adopt the attitudes, and behaviours that mirror bigotry.

I haven't felt the need to construct a lexicon of derisive terminology to discuss this topic. That does not seem to be the case for others.

Companies and technologies are not people. They are not analogous.

I do not think they need to be people in order to oberve that people behave in a manner that resembles bigotry.

Is it your contention that the the epithets and derision do take the same form as bigotry but do not qualify as such because the targets of derision lack personhood?

I am doubtful that any existing AI has the capacity for personhood, but I think you are taking the position that no AI could ever be a person.

Would it be sufficient for some people to believe they are people for it to be bigotry? Or would you consider the behaviour to not be bigoted if they did not consider the target to be a person?

Reply on news.ycombinator.com