@alexandra I'm trying to wrap my head around this argument (if you feel it's inappropriate, please let me know and I'll stop, don't block me!)
I have serious ethical misgivings about people who drive cars, for the very same reasons, it's unjust, violent, etc. and advocate for a fairer city. Based on this reasoning, would I also need to reject contributions from people who drive cars?
Nov 8, 2026, 2:24 AM@mishari @alexandra funny you mention a fossil fuel driven industry, given the enviromental harm and health concerns known for more than 70 years ago, and we still won't get any "regulations" as promised, and the industry being an active lobbist for the governments to stop said regulations. And even with those regulations, it is still an exploitative and killer business worldwide (petro-dollar, middle east interventions by the west, wars, etc.).
A lot of the pro AI discourse is about how "genAI is not that bad, it just need to be regulated". And we know for sure that wont happen anytime soon.
@xaq another excellent example, does one then reject products and services powered by fossil fuels? @alexandra
Nov 8, 2026, 2:56 AM@mishari @alexandra I think it is indeed a capitalism struggle, but not comparable.
An example I'm thinking about is this: You can code without LLMs, just as you can also transport food on your back by walking to markets so people can buy it in order to eat. Both would take some time, but food bc is quickly fungible, code will not rot, food will. Sometimes, time matters. And in this case we can compare what it is to use a problematic system to fullfill a fisiological need (transport food so people can eat, nourishment, health) with using another problematic system to "shortcut" workload or tools. This got me thinking about any real life example in which oneself really need to rush things that it is needed to code using LLMs or agentic AI bc we can't just wait to do it ourselves as the world did before 2018? Or is it just hype?
Of course we can boycott the fossil fuel industry, people can commute to work, use the public trasnport system, make chages in its routines, and every small scale transirion that it os available to their reality and possibilities. And still the way our society has been built will still rely on fossil fuel to power trucks, buses, planes, ships, etc. to mantain the many industries that depend on them.
But then, I contemplate genAI as something that it's just getting started (I'm not talking about machine learning), it hasn't been implemented as the foundations of tons of industries (besides the hype).
We are still on time to be more cautious about these things, and to be more critical 🫂
@xaq @mishari @alexandra I would also consider "there's no ethical consumption under capitalism"
Just because someone drives a car doesn't mean they're strictly evil, and if your job requires LLM use and you use it for your job then 🤷
But the equivalent question would be "should I reject contributions from someone just because they buy a 4 gallon per mile car from Donald Trump?"
And... yeah that sounds like a terrible person that we would not want un-repentant in our communities.
Nov 8, 2026, 5:11 AM@ketmorco @mishari @alexandra a cop would also state that they are forced to do profiling or k*lling innocent ppl "bc they are forced to do it on their job", but you really arent forced to do it, you are being actively and purposelessnessly coerced to do it. You can always choose not to comply and quit. But you actively choose not to bc many reasons. That doesn't mean you're not causing the harm and concecuences that are a byproduct of your actions, even if you're "just working", "just buying food" or "just meaning to thrive". Part of being human person is that eventually you're going to harm other people at some point. We all do it, it happens. Causing harm doesn't make you a bad person, it is how you respond to that, how you tske responsability for that. This idea that only bad people do bad things makes us unaware that we're causing harm to to other people bc we believe we're not bad people so we couldn't have hurt somebody else.
Going back to the trolley problem. We can choose to switch rails to save 5 ppl instead of 1, but we can also question: can we derail the trolley and not roll over anybody? Why are we the ones who drive the trolley? Were we qualified to do it? Who tied those ppl to the rails? Who taught that someone how to make knots and kidnap ppl? Why is that someone doing it? Why the trolley company didn't apply safety measures in order to prevent this plausible situation? By making these questions and observations maybe we do have more alternative actions to do in order to aproach the proposed conveniently simplistic problem on which we have to decide. Maybe it's not for us to decide on whos life is worth taking for saving others in exchange. Maybe we could be working in another job as a barber, a chef, a waiter instead of working as a trolley chofer. Maybe we should consider it if that same situation or the like had already happened before.
The "no ethical consumption under capitalism" situation is a lot more nuanced than that. You can comply and wash away your guilt and counciousness by convincing yourself that you did what you were supposed and abled to do, but if you really weren't aware of the unethical part of doing what you did, then you probably wouldn't find yourself in a struggle for what was done and its consecuences.
@xaq @ketmorco @mishari @alexandra this isn't corporate social media. You do not need to censor the word "killing".
Nov 8, 2026, 8:44 AM@mishari @xaq @alexandra
Noone really addressed Eli's point about fragmentation. Slavery is a clear case of evil, but what about other cases?
Let's take another hypothetical case where emacs fragments once again between Emacs-US and Emacs-China. Both consider the other evil. What do we do then?
And I also think that "evil" is the wrong question. Does it harm development?
If a developer is abusive to others? yes. If a developer floods the bug tracker with slop: yes.
@tzafrir @mishari @xaq @alexandra
> Both consider the other evil. What do we do then?
Research what they are doing and what they stand for. Just like with any other thing.
> And I also think that "evil" is the wrong question. Does it harm development?
Nah, full stop. There are more important things than some silly software development.
Nov 8, 2026, 12:45 PM@tragivictoria @mishari @xaq @alexandra
Surely there are more important questions. And there are plenty of other venues to discuss them.
The mailing list emacs-devel, however, is about software development. A very specific software (even though some people claim it includes all you ever need from a computer).
@tzafrir @mishari @xaq @alexandra
> The mailing list emacs-devel, however, is about software development.
And one of the maintainers said there that slave labor is OK. If you do not see a problem here, then there's nothing to discuss.
Nov 8, 2026, 12:55 PM@tragivictoria @mishari @xaq @alexandra
Not really. This was an extreme hypothetical example raised by someone else (Brian). Eli's response was: we can't start checking moral compass of every contribution. It's impractical and murky.
Do we make every contributor sign the JSON license[1]?
If you have any specific objections, please make them. "Evil" is not a useful one.
@tzafrir @tragivictoria @mishari @xaq @alexandra no, we make contributors sign CoCs. And yes, that’ll cause a split. Good.
Nov 8, 2026, 1:54 PM@chucker @tragivictoria @mishari @xaq @alexandra
A CoC is very practical and very actionable. I have no idea who Jia Tan, or Jane Doe, is. They might actually be a slave trader or a botnet operator. I can't tell.
But if Jia Tan starts misbehaving around my community, that is something I know very well and should react to.
A CoC is not about who you are or what you think. It is about what you do in our community (mailing list, conference, whatever).
Nov 8, 2026, 4:02 PM@tzafrir @tragivictoria @mishari @xaq @alexandra “if we reject one evil thing we’ll need to reject other evil things too, WHERE WILL IT END” is not the massive slam dunk you think it is
@mishari @xaq @alexandra Here's the thing, as of now, oil rules. But that wasn't always the case. Even if we couldn't avoid it, we could've enacted more fair profit and cost split.
With LLMs we are before they rule. The bet is that the first one to have LLM so powerful and widely used, that it gets industry addicted to it, will gain near unlimited rent from all around the world. It's oil 2.0 but worse. Oil gave physical capabilities. Here the fight is for competencies. For being able to effectively withhold the labor, hence having political power to fight for rights and retain them.
If ownership of such tech is private, then power shifts to the private hands. Until this changes, resisting adoption is a form of protest.
Nov 8, 2026, 4:51 AM@mishari
the example involved slaves being forced to produce code, which the slaver would submit as their own, and the project would know about all of it, to which this emacs maintainer said, "why yes of course!"
cars aren't producing code, so it's not directly comparable.
you also have to think about things like intent to harm others (importantly, as distinguished from intent to protect others from harm), actual harm caused, individual impact of the behavior, widespread societal / cultural entrenchment of the behavior, and whether acceptance of the behavior (or person doing it and refusing to stop) could increase its occurrence or cause it to spread.
slavery should be an obvious "no, never"! cars, not so much. LLMs are not nearly as obvious, but their use does cause harm, and accepting LLM code does cause it to spread and worsen the problem. whether someone uses LLMs in outside projects is less an issue, imo (though still not great). accepting or rejecting code from someone who drives a car? zero impact whatsoever on the problem and it's not related at all.
Nov 8, 2026, 9:53 AM@tmw @mishari "protecting others from harm" is just a way to justify yourself causing harm and to not feel bad about it; its a way to dismiss it as being super "nessecary" furthermore your intentions dont fucking matter at all what matters is what actually fucking happens; slave owners can claim its ""for their own good"" because their incapable of doing anything good for themselves and its actually helping eveyone becuase the blacks are all harmful and bad and keeps them harming people .... and other racist bullshit all the same too, just as a cop will insist brutalizing everyone around them is for your safety actually and so on .. it does not infact make it magically fine.
Nov 8, 2026, 4:57 AM@mishari slopping is a choice, driving cars often isn't.
Nov 8, 2026, 7:55 AM@mishari Some thoughts. Reasoning about it step by step, ideally yes, especially if that would have any practical impact.
In practice, banning LLM contributions at this moment in time is a much stronger political position than not accepting contributions from people who use cars - and also, everyone is still able to make contributions without LLMs, unlike with cars which for some countries are already a systemic problem.
LLMs will become an entrenched systemic problem if we don't work against that, but they aren't yet.
Although, banning people who use cars certainly seem like an interesting radical position. We certainly need movements like that.
Nov 8, 2026, 8:54 AMNov 8, 2026, 10:10 AM@mishari @alexandra whats being suggested is actually blocking contributions known to be made by slave labour (or otherwise, made by an LLM) not nessecarily banning anyone who uses LLMs it is litterally just a "dont use them here" ..
so the more accurate comparison would be if somehow the car was involved in the contribution to emacs somehow .. there is probably a argument to ban slave owners and maybe even anyone who uses an LLM from contributing but .. that like isnt even whats being suggested .. its way more leniant than that-- and their still throwing a fit from it.
Nov 8, 2026, 4:00 PMyet you participate in driving. curious!
@jacques @mishari mister gotcha strikes again https://thenib.com/mister-gotcha/