[rendang]

One could probably think of dozens of reasonable arguments for avoiding LLM use, but this one is awful. If LLMs actually are able to get more work done with fewer people aka "firing people" that would be wonderful for humankind. If you disagree and like getting less work done with more people, you are welcome to forego tractors, dishwashers, the steam engine, and all the rest.

[UqWBcuFx6NV4r]

Yeah. This has been an interesting cultural shift, especially with “the kids”. I’ve had at least a few people passionately tell me that using (non-generative, non-LLM) AI to assist with social network content moderation, is unethical, because it takes away jobs from people. Mind you, these are jobs in which people are exposed to CSAM, gore, etc. A fact that does not dissuade people of this view. There are certainly some sensible arguments against using “AI” for content moderation. This is not one of them.

It’s really intriguing how an increasingly popular view of what’s “ethical” is anything that doesn’t stand in the way of the ‘proletariat’ getting their bag, and anything that protects content creators’ intellectual property rights, with no real interest in the greater good.

Such a dramatic shift from the music piracy generation a mere decade or two ago.

It’s especially intriguing as a non-American.

Again, as you say, many sensible arguments against AI, but for some people it really takes a backseat to “they took our jerbs!”

[majormajor]

I can't speak to outside the US, but here companies have gotten much more worker-hostile in the last 30 years and generally the economy has not delivered a bunch of wonderful new jobs to replace the ones that the information age already eliminated (let alone the ones that people are trying to eliminate now). A lot of new job growth is in relatively lower-paying and lower-stability roles.

Forty years ago I would've had a personal secretary for my engineering job, and most likely a private office. Now I get to manage more things myself in addition to being expected to be online 24x7 - so I'm not even convinced that eliminating those jobs improve things for the people who now get to self-serve instead of being more directly assisted.

[poulpy123]

Last time I checked, most people needed a "jerb" to buy food, have a shelter or provide their children. So when the promise is to lose this "jerb" they are fully in the right to be scared.

[andrepd]

They took our jerbs is a perfectly valid argument for people which face ruin without a jerb.

Capitalism is not prepared nor willing to retrain people, drastically lower the workweek, or bring about a UBI sourced from the value of the commons. So indeed, if the promises of AI hold true, a catastrophe is incoming. Fortunately for us, the promises of AI CEOs are unlikely to be true.

[TeMPOraL]

Yup. The problem was never with the technology replacing work, it was always with the social aspect of deploying it, that ends up pulling the rug under people whose livelihood depend on exchanging labor for money.

The luddites didn't destroy automatic looms because they hated technology; they did it because losing their jobs and seeing their whole occupation disappear ruined their lives and lives of their families.

The problem to fix isn't automation, but preventing it from destroying people's lives at scale.

[expedition32]

You don't ban new technology you regulate it.

That is what happened with the 19th century factories.

[barishnamazov]

I personally wish for the time when AI can replace everything I can do (at least in my current field). I'm not sure how exactly I'll feel about it then, but it'd be a technological advancement I'd enjoy seeing in my lifetime.

One question perhaps is, even if AI can do everything I can do (i.e., has the skills for it), will it do everything I do? I'm sure there are many people in the world with the superset of my skills, yet I bet there are some things only I'm doing, and I don't think a really smart AI will change that.

[poulpy123]

I would like for an AI to do my work, unfortunately I have to buy food and pay my rent.

The Industrial Revolution caused a great deal of damage. It was a net positive in the long term because new jobs were created to replace those that were lost, but it took decades and enormous violence. Now, the promise of AI is that it will be more efficient than any human being. If this becomes a reality, there will be, by definition, no new jobs created for the people replaced by AI.

[andrepd]

Would being the operating word here. In a capitalist economy with wage labour, it is a catastrophe.

[goatlover]

Does this argument still work if LLMs end up increasing unemployment and making it a lot harder for graduates to find good jobs? Who is it good for in that case, the shareholders? It's nice if humans can always create more jobs, but that's not what the tech bros are promising investors. They're making claims about how AI is going to seriously reduce the need for human labor. Programming, writing and art are just the starting ground for what's coming, if their predictions are anywhere close to being correct.

[rendang]

Because consumer demand is infinite, the only way to majorly and permanently increase unemployment is to completely automate all labor or maybe almost all. We have been automating away jobs for hundreds of years and unemployment is still ~4%