Post by @publicvoit@graz.social

RE: https://mastodon.social/@meejah/117595425483136501

I'm not going to sign #HumanEmacs because I *want* discussions about LLMs effectiveness in particular with open weight models, for example.

You can't put the genie back in the bottle.

LLMs are here to stay. Ignoring their impact without any discussion as proposed on https://human-emacs.org/ is the wrong approach in my opinion. Blocking is not the best thing here, discussing the terms would be better.

If not, what's next? Block all people who use hardware where its vendor has some connections to unfair workforce policies? Block all #FOSS contributors whose electricity supplier as a certain amount of nuclear energy?

It's never black or white.

I think we've got enough experience with #cancelculture, over-stressed #policitcalcorrectness and similar movements already.

I acknowledge the good spirit behind it, I disagree on the chosen method.

YMMV.

#AI #LLM #openweight #Emacs

#policitcalcorrectness #cancelculture #humanemacs #openweight #emacs #llm #foss #ai

0 Boosts

0 Likes

David Pape

@zyzzyxdonta@social.tchncs.de 2 hr. ago

@publicvoit to me, "cancel culture" is a right-wing inflammatory term trying to discredit valid critique as an attempt to silence people. The text on the Human Emacs website does not call for any particular person to be chastised. Neither does it make any threats or implications about people who have other opinions. They are simply not mentioned at all. Please point out any snippet in the text that you would interpret differently.

This federated comment is not yet available here.

Karl Voit :emacs: :orgmode:

@publicvoit@graz.social 2 hr. ago

@js Do I? 🤔

Maya Granade

@kevingranade@mastodon.gamedev.place 3 hr. ago

@publicvoit good, we don't want you.

ben

@benjamineskola@hachyderm.io 4 hr. ago

@publicvoit

> LLMs are here to stay.

so is asbestos but we don't have to make the problem worse.

Karl Voit :emacs: :orgmode:

@publicvoit@graz.social 2 hr. ago

@benjamineskola Well even asbestos is an existing reality.

So we do have to discuss how to deal with asbestos and quite frankly where I live, this is an ongoing discussion.

The current letter explicitly wants to avoid any discussion. Which I do find a pity because without a real discussion, all the negative aspects will get continued without any impact.

Alex Schroeder

@alex@social.alexschroeder.ch 5 hr. ago

@publicvoit I don't share your interpretation at all. I think people discuss LLMs everywhere these days. Once an individual has reached a conclusion and wants to sign up for Human Emacs, then there is no point to discuss LLMs with the other signatories.

I also don't understand the qualifications you make regarding cancel culture and over-stressed political correctness. To me, not wanting to collaborate with people who have terrible opinions is the most natural thing to do. Sometimes I have to. Sometimes its expedient not to know. But other than that, having values and living up to them is what I expect of everybody and specifically of those I am going to collaborate with.

Karl Voit :emacs: :orgmode:

@publicvoit@graz.social 5 hr. ago

@alex

> I don't share your interpretation at all.

Which interpretation? That they don't want to discuss? Made very clear on https://human-emacs.org/

> people who have terrible opinions

That's the point. Who defines what "terrible" means? I don't see any direct connection between people who want to discuss LLMs and "terrible opinion".

Human Emacs

Alex Schroeder

@alex@social.alexschroeder.ch 4 hr. ago

@publicvoit I don't share your interpretation of the text on the web page. You say it's very clear on the page and yet we disagree.

The Human Emacs project is for the people who have already discussed LLMs and have reached the conclusion that they don't want to use LLMs, nor their products. Therefore, do not sign on and then discuss the LLMs. Sign up if you are already convinced.

If you want to discuss LLMs, no problem. Don't sign up. But Human Emacs is not "ignoring their impact without any discussion". We have been discussing LLMs for years. It's all around us. It's inescapable. Surely there must be a place somewhere for the people who have reached the conclusion that they don't want LLMs in their Emacs.

As for "terrible": Of course it's a value judgement. But it being a value judgement does not result in inability to judge. If you don't see how LLMs are terrible, then you have reached a different conclusion than other people. Which is fine. But these other people still get to act upon the conclusion they reached, obviously. People disagreeing with each other and acting upon this disagreement is all around us.

I understand that if you don't see how LLMs are terrible then everything that follows makes no sense. As for myself, I've kept a long list of bookmarks over the last few years. Every now and then I still add new links. I think it would be fair to say that clearly there are some people who think that LLMs are terrible, even if you don't share this opinion.

Bot Check

Karl Voit :emacs: :orgmode:

@publicvoit@graz.social 2 hr. ago

@alex
> I don't share your interpretation of the text on the web page. You say it's very clear on the page and yet we disagree.

I wrote that they don't want discussions.
They wrote: "We are not here to discuss ..."

I am really puzzled how we could disagree here. 🤷

tusharhero

@tusharhero@mathstodon.xyz 5 hr. ago

@publicvoit @alex I define what terrible means to me. Simple as that. What's so complicated about that?

Karl Voit :emacs: :orgmode:

@publicvoit@graz.social 2 hr. ago

@tusharhero @alex I do think that this is terrible. 😀

1+ more replies

josemanuel

@josemanuel@social.sdfeu.org 5 hr. ago

@publicvoit I don't think they're saying that at all. My understanding is that they point out that LLMs are a net negative technology and that ethical software projects should not use them. According to Eli, Emacs doesn't care about ethics, so a fork seems the way to go.

I don't think cancel culture or any of the other terms your threw in your post apply here. It's just a difference in the way free software is understood by each side.

Karl Voit :emacs: :orgmode:

@publicvoit@graz.social 5 hr. ago

@josemanuel Well: https://human-emacs.org/

> We are not here to discuss whether LLMs are effective at what they are claimed to be able to do

> We are not here to discuss "open weight" models

What did I mention where you think it's not part of this movement?

Human Emacs

josemanuel

@josemanuel@social.sdfeu.org 10 min. ago

@publicvoit I didn't say that. What I did say is that they consider these technologies a net negative in ethical, social and environmental terms. It makes sense to me, when we're talking ethics and principles, to not have them be open to negotiation, but that is not the same as cancel culture or political correctness.

Are they taking the “my way or the highway” route? Probably, but it's them who will take the highway if their demands are not met. They're just asking for a reason to stay.

arpunk

@arpunk@chaos.social 5 hr. ago

@publicvoit "Rejecting your PR because while your 2-line typo fix is clean, your diet wasn't vegan during the commit, making this patch complicit in global injustice and therefore will be auto-closed to maintain project karma."