Saw some article about "quitting LLMs for climate" and.. uhg. Can we please drop the "personal responsibility" bs? Your choice, as an individual, to use or not use this shit has no effect on the climate.

Where your individual use of LLMs *does* have harm is rotting your brain, destroying your credibility, exposing you and your contacts' private conversations and photos to parties who wish us harm, wasting your peers' time, exposing people who use the things you made to legal risk, etm etm etm.

Edit: To be clear, I'm perfectly happy with people refraining from using LLMs for whatever reason works for them! This post was more about media/mass-communication narratives and how they risk implying that there's value in "AI" and that refraining from using it is a sacrifice rather than being self-care and community care.

@dalias I'm starting to suspect that the detachment and psychosis LLMs cause might just be a feature to many affected

@dalias Next you're going to tell me that not buying fast fashion has no effect on how much is wasted! 🙃

via @breadandcircuses

https://climatejustice.social/@breadandcircuses/112524382847749783

Bread and Circuses (@[email protected])

Attached: 1 image I did a post about this from a different source back in January. And the problem is not going away. In fact, it’s only getting worse as capitalist industry and advertising continues to cram 'fast fashion' down our throats. _______________________________ The Atacama Desert in Chile has attained a distinction as one of the world’s fastest growing dumps of discarded clothes, thanks to the rapid mass production of inexpensive attire known as fast fashion. The phenomenon has created so much waste that the UN calls it “an environmental and social emergency.” The numbers tell the tale. Between 2000 and 2014, clothing production doubled and consumers began buying 60% more clothes and wearing them for half as long as they once did. Three-fifths of all clothing is estimated to end up in landfills or incinerators within a year of production. That can translate to a truckload of used clothing dumped or burned every second. In northern Chile, colossal piles of discarded clothes, with labels from all over the world, stretch as far as the eye can see. In one ravine on the outskirts of Alto Hospicio, a pile of jeans and suit jackets, bleached by the harsh sun, rises above a mound of fake-fur coats and dress shirts, some still bearing price tags. _______________________________ FULL STORY -- https://archive.ph/1z2Zf #Environment #Climate #Capitalism #BusinessAsUsual

Climate Justice Social
@catsalad @dalias @[email protected] We should make robots that crush the garbage into compacted cubes and- WAIT HOLD ON

@catsalad @dalias @breadandcircuses I have many thoughts about pushing personal responsibility at people for things like climate change.

However -- I don't think it's wrong to make personal choices based on such things. Especially with LLMs because people's choice to use them is entirely optional and they do have negative external effects. If more people choose not to use them, eventually the big tech push to use them is going to fail. If more people choose to use them, we're stuck with them. It does boil down to personal choice here in a lot of ways.

Also - they're choosing not to use LLMs, so ... it's a win. Take the W.

@jzb @catsalad @breadandcircuses My problem isn't with them making a personal choice that reflects values and makes a statement. It's with framing it as a *sacrifice* for the planet. Which implies the thing they're "giving up" is good, is something that would give them an advantage if they took it, and that they're doing something noble putting themselves at a disadvantage by not doing so.
@dalias @jzb @catsalad @breadandcircuses And what if that is *exactly* how they believe it is? That may indeed be their personal viewpoint.
@neal @jzb @catsalad @breadandcircuses Then it's a narrative that's actually promoting "AI".

@dalias @neal @jzb @catsalad @breadandcircuses To me, this framing is intended to meet everyday people where they are. LLMs are fun to poke and get slop out of, for a while. It's indistinguishable from magic to a non-techie!

So by saying, "Hey, look, I know ChatGPT is fun, but the planet is more important," one may be giving a hollow compliment to the tech, but they immediately quash it with a reality check. This is getting one's foot in the door by being relatable, not LLM-aggrandizement.

@GrandTheftUrkel @dalias @jzb @catsalad @breadandcircuses Also, as a general rule, your approach of bashing everyone for this *isn't working*. It doesn't help, it doesn't make anyone reconsider *anything*, and you are just pushing people to reinforce their position regardless of what it is.

You need to get them to care for your position, and your approach makes them do the exact opposite.

@neal @GrandTheftUrkel @dalias @jzb @breadandcircuses So instead of bash, use zsh⁠?

@catsalad Zshushing people is a different way of bashing people, so it would likely result in the same outcome that @neal warns about?

Now, if the discussion could be gently turned in the direction of teaching people to fish...

@neal @GrandTheftUrkel @dalias @jzb @breadandcircuses

@neal @GrandTheftUrkel @dalias @catsalad @breadandcircuses One of the things I've been thinking about with regards to LLMs is the similarity with Free Software.

That is: Free Software approaches software as an ethical thing. I agree. *But* it's a non-starter if people don't have an ethical framework in their thinking about software. And most people don't -- so for the Free Software arguments to work, the first step is convincing people that there is an ethical dimension to software *at all* before persuading them to follow a specific ethical framework.

Likewise: ethical arguments around LLMs only work if people have an ethical framework that those arguments resonate in.

Backing up, if people feel that they're being noble and making a sacrifice by not adopting/pushing LLMs... I'm not going to argue with them on that point.

I'd rather be successful than right but unsuccessful. 🙂

@neal @GrandTheftUrkel @dalias @jzb @catsalad @breadandcircuses did we do that when we made racism and homophobia socially uncool?
@outfrost @GrandTheftUrkel @dalias @jzb @catsalad @breadandcircuses Yes. Connecting it to principles they do care about and making the incongruity aware makes it much easier to persuade and convince people.

@outfrost @neal @dalias @jzb @catsalad @breadandcircuses This is a false equivalence lol. But I agree with Neal's point; if you want to change hearts and minds, you have to meet people where they are. To get someone to leave their comfort zone, you must sometimes leave yours. And you don't need to agree with a position to empathize with it.

Expecting someone to magically snap to your way of thinking, simply by being outwardly disgusted enough by their position, is an exercise in masturbation.

@GrandTheftUrkel if you're trying to change my mind then this reads pretty ironic

so, for the "hearts and minds" thing to work, the recipient must already care about either you personally, or keeping their views consistent. a pretty tiny number of people qualifies, unless you're a celeb. the rest don't really care about your opinion until there are direct consequences to ignoring it.

the political climate has changed dramatically in the last 10-15 years.

@GrandTheftUrkel @neal @jzb @catsalad @breadandcircuses I guess it's a matter of whether you see it like "quitting smoking" (something that's also beneficial to your own health) or "giving up [something] for lent". But a lot of the takes I've seen feel like the latter.

@dalias

A lot of people see value in it. "Here's the bad things you also agree are bad" is a much better argument than "but you're wrong though".

Tbh i think you're more focused on being right than on changing people's minds. Imo, being right is irrelevant if people ignore you.

@neal @jzb @catsalad @breadandcircuses

@catsalad

Software developer sits here bemused, wearing tee shirt, jeans and trainers that all have holes worn into them. You mean people buy new clothes from choice?

@dalias @breadandcircuses

@dalias good take! i tend to categorically differentiate personal usage vs public advocacy, as it's a common tactic to describe "having morals" in terms of censorship and to ablate power distinctions. when you post publicly to describe personal use of LLMs, you are in fact endorsing their use. you have identified an additional subtlety here, wherein the public discussion is conflated with personal use in a way that confuses the audience further
@dalias nah, I never used it because of climate care and not gonna pretend it have no effect.
Better hurry up bc your point is a fast moving object, and is moving away.
@dalias Yea, same with my private jet. Why should I not use it if all my friends still use it. Fuck the climate.
@burningTyger That's not related because ordinary ppl don't have private jets. There is no private jet industry pushing billions of ppl to fly private jets. It is not a harm that's the aggregate of billions of microharms that individually mean nothing and where only fixing the pressure to do it can solve anything.
@dalias People in Nigeria don't have a computer either. At least the people I know and who work in higher education.
@burningTyger What does that have to do with anything?
@dalias Things that "ordinary people" have or have not. Things like private jets or computers. Every choice is embedded into your socioeconomical context and should be considered meaningful.
@burningTyger In "ordinary people have" the point is not whether everyone has it. The entire point is that it doesn't matter if one person, or even a large class of people, doesn't, because the harm comes from the gigantic scale, outside the control of anyone who doesn't, of people who do. And that is not a result of their aggregate individual choices (which individually possess "free will" but at scale are clearly modeled by predictable rules) but of the pressures imposed, by entities in possession of power, on populations.
@dalias now you lost me to your argument :)
Can we agree that LLMs are bad?

@burningTyger I believe cars, at the scale they're produced, owned and used today, are bad.

I also own and drive a car (while living in a densely populated European city).

It's the system that's broken, not the decisions individuals make in order to make their lives easier within said system. A few people choosing to not own a car doesn't fix the climate, health, safety, or community impact of cars in general.

I believe this is roughly what @dalias is saying wrt llms.

@outfrost @dalias I know and it frustrates me because I do think my actions have impact. Who's going to start making a difference. While I personally have some disagreements with Greta Thunberg's activities she started something small with global impact. Who know who is next? Might be me, might be you. It's an ethical decision for me. I try not to judge other people for their decisions but I want to make it better for everybody so I try to do as much as I can on my own.

@burningTyger @outfrost Your actions in systemic things can make a difference when they're part of an organized concerted effort to produce change and take into account the power systems involved.

When they don't make a difference is when you just think "well I'm not doing that, so I've done my part and the other people who are doing the thing are the problem" rather than identifying the power relationships that lead to the thing being done at population-scale.

@dalias @burningTyger yeah. Beverage corporations taught people to "recycle" plastic bottles not because that fixes the plastic problem, but because they want you to think it's your responsibility, and that if only everyone sorted their rubbish properly, there would be no plastic problem. Because they want to keep selling overpriced beverages in cheap plastic bottles with no consequences.

@burningTyger That said, I do always appreciate strongly principled personal choices! What you do in this regard is definitely good. I guess we just need to not turn this issue into another plastic bottle.

@dalias

@burningTyger @dalias
My current job involves coding on things I think shouldn't exist. (And I don't get much choice in the market, really, beyond not touching slophouses.)

If, hypothetically, LLMs only had a climate, hardware availability, and labour market costs, I could see myself using an LLM sometimes at this job. I need to pay my bills, I don't care about the product, in fact lowering the code quality could be a benefit.

But in reality the evil of LLMs goes so, so much deeper than that.

@burningTyger @dalias That is of course _not_ to say i'd be happy with LLMs in the described hypothetical. But it does assume them being built like an actual tool, through genuine means and with correct intentions, which couldn't be further from the reality we face.

And I think there's value in teaching people that nuclear weapons are not "just tools".

@dalias @outfrost Now you're talking Niklas Luhmann and the resonance between systems. You have to find a way in your system that creates vibrations in the other system so that they oscillate with your system. I'm bad at explaining this. But you have to make your idea their idea. This works pretty well e.g. in making green energy popular with conservative folks if they see that it makes them save a lot of money. The tricky part is finding the sweet spot that triggers the other system.

@dalias AI is like a brand new oil industry, and we're definitely pushing to get off of oil.

Sure, it's not our fault the industry exists, but we can be responsible in our choice to use it or not.

There are so many reasons to not use AI, but to me climate is the first and most important reason.

I saw your later more nuanced point about sacrifice framing being problematic because it implies AI is valuable. But many people's experience with AI is that they will first be exposed to (false) claims of its value, before learning of its harms. So they likely are going to have this "I thought it was valuable but I'm quitting because of X harm" moment, and I don't think that's something to be upset about. Later they will learn that what they are giving up was never really valuable at all.

@dalias Well said!

For the harm column, you can also add in “giving money to the people who are funding fascism/amassing personal power through obscene amounts of wealth”. No one person stops wealth inequality or fascism by themselves and their own choices: that takes mass, organized and focused movements, not merely many individual people, that have the power to resist.

@dalias i think it's perfectly rational to avoid LLMs because they damage the climate. you may not want to get that blood on your hands; i see that.

of course it's silly to think that one person doing that will make a difference.

but it's also silly to think that *everyone* doing it would not. i hope anyone that thinks this, isn't voting? because, same deal.

@dalias it is shit for all these reasons and the ecological reasons and everyone must abstain from using "AI" also for ecological reasons.
@mirabilos What I'm saying is that a bunch of people individually abstaining for ecological reasons isn't going to solve the ecological problem. Me not talking to ChatGPT does nothing if every company providing services I need has replaced their customer support and operations staff with chatbots. We need to bring down the people at the top pushing "AI", regulate/tax/ban the "AI" data centers out of existence, etc. Not just try to convince the few people willing to make sacrifices for ecological reasons to stop rotting their own brains on 🐈💨

@dalias yes-ish, until the “bunch” gets large.

But for a personal reason, refusal to use planet-burning things is still a good thing (also relevant for… most religions, I’d wager).

@mirabilos @dalias I'd put it as not just abstaining but also walking away (as much as possible I guess…) from anyone and anything which uses it, which would mean having to seek alternatives and sometimes build them.