A tech news sub for communists
LLMs really might displace many software developers. That’s not a high horse we get to ride. Our jobs are just as much in tech’s line of fire as everybody else’s have been for the last 3 decades. We’re not East Coast dockworkers; we won’t stop progress on our own
why did I do computer science god I fucking hate every person in this field it's amazing how much of an idiot everyone is.
You can tell the ones that got A's in their comp sci classes and C's in their core/non-major classes by how bloodthirsty they are.
Me, the enlightened centrist, just got C's in everything
Writer is the type of guy to only fail his ethics courses
The argument that workers should capture AI instead of the ruling class is interesting, but let me ask you.
Has there been a single technology entirely captured and for the workers in history, ever? Has not every piece of technology been used primarily by the working class, yes, but the direction it develops and what value it produces is decided by the ruling class? Always has been unless we can remove them from controlling the mode of production..
I think China is an interesting example of this, where the worker's party controls the majority of the economy and wouldn't let a program like DeepSeek threaten to unemploy half of it's economy (America does probably have a larger segment dedicated to programming, though, silicon valley and all). Even then, the average worker there has more safety nets.
The threat I see is the dominance of AI services provided by an oligarchy of tech companies. Like Google dominance of search. It's a service that they own.
Thankfully China is a source of alternative AI services AND open source models. The bonus is that Chinese companies like Huawei are also an alternative source of GPUs. This allows you to run your own AI models so you don't necessarily need their services.
You're thinking of class war. There's only one proven way to win that war: The working class rises up, kill some MFers and takes over.
I mean, technology will be used to oppress workers under capitalism. That is why Marxists fundamentally reject capitalist relations. However, given that people in the west do live under capitalism currently, the question has to be asked whether this technology should be developed in the open and driven by community or owned solely by corporations. This is literally the question of workers owning their own tools.
If people can build it, it can serve the people. Think of open-weights LLMs. If we got a couple of 32B models that score as high as GPT-4o and Claude-3.5, why not use them? It can be run on mid-high end hardware. There are developers out there doing a good job. It doesn't need to be a datacenter/big tech company centered scenario.
Thanks for sharing these AI posts.
Paid employment could mean retraining under socialism. Remember communism is moneyless, stateless and classless. The aim of society is the socialisation of all labour to free up time to do more leisure including art. People will still want art from humans without AI but there’s a difference between that and the preservation of regression through ludditism to maintain less productive paid labour.
Equating anti-capitalism to anti-corporatism, the appeal to ludditism, the defense of proprietorship, or the appeal to metaphysical creativity is not going to cut it, and that is a low bar to clear for marxists.
My party is trying their best to understand and implement AI and it's causing some friction within the party. The official stance that is now adopted is the one of 'we need to understand it and use it to our advantage' and 'we need to prevent AI being solely a thing of the ruling class' and to me that makes sense. I wasn't around at the time but I imagine it was the same with the coming of the internet some decades ago and we can see how that ended. I hope socialist orgs don't miss the boat this time.
I think that's precisely the correct stance. As materialists we have to acknowledge that this technology exists, and that it's not going away. The focus has to be on who will control this tech and how it will be developed going forward.
I find the tone kind of slapdash. Feel like the author could have condensed it to a small post about using AI agents in certain contexts, as that seems to be the crux of their argument for usefulness in programming.
I do think they have a valid point about some in tech acting squeamish about automation when their whole thing has been automation from day one. Though I also think the idea of AI doing "junior developer" level of work is going to backfire massively on the industry. Seniors start out as juniors and AI is not going to progress fast enough to replace seniors probably within decades (I could see it replacing some seniors, but not on the level of trust and competency that would allow it to replace all of them). But AI could replace a lot of juniors and effectively lock the field into a trajectory of aging itself out of existence, due to it being too hard for enough humans to get the needed experience to take over the senior roles.
Edit: I mean, it's already the case that dated systems sometimes use languages nobody is learning anymore. That kind of thing could get much worse.
The developer pipeline is the big question here. My experience using these tools is that you absolutely have to know what you're doing in order to evaluate the code LLMs produce. Right now we have a big pool of senior developers who can wrangle these tools productively and produce good code using them because they understand what the proper solution should look like. However, if new developers start out using these tools directly, without building prior experience by hand, then it might be a lot harder for them to build such intuition for problem solving.
Do you like fine Japanese woodworking? All hand tools and sashimono joinery?
this should sell it, why would anyone want something more expensive just because it was hand made instead of mass produced.
Turns out my assumptions about how LLM-assisted programming works were completely wrong or outdated. This new way sounds super efficient.
Every six months the tone of these "why won't you use my hallucinating slop generator" get more and more shrill.
I think his point that you basically give a slop generator a fitness function in the form of tests, compilation scripts, and static analysis thresholds, was pretty good. I never really thought of forcing the slop generator to generate slop randomly until it passes tests. That's a pretty interesting idea. Wasteful for sure, but I can see it saving someone a lot of time.
you basically give a slop generator a fitness function in the form of tests, compilation scripts, and static analysis thresholds, was pretty good.
forcing the slop generator to generate slop randomly until it passes tests.
I have to chuckle at this because it's practically the same way that you have to manage junior engineers, sometimes.
It really shows how barely "good enough" is killing off all the junior engineers, and once I die, who's going to replace me?
This is absolutely the crisis of aging hitting the software engineering labor pool hard. There are other industries where 60% or more of the trained people are retiring in 5 years. Software is now on the fast track to get there as well.
This is a great point. I think what is most jarring to me is the speed at which this is happening. I may be wrong but it felt like those other industries, it took at least a couple decades for it to happen, and it feels like tech is doing it in a matter of months?
I'd much rather the slop generator wastes its time doing these repetitive and boring tasks so I can spend my time doing something more interesting.
wastes its time doing these repetitive and boring tasks
To me, this is sort of a code smell. I'm not going to say that every single bit of work that I have done is unique and engaging, but I think that if a lot of code being written is boring and repetitive, it's probably not engineered correctly.
It's easy for me to be flippant and say this and you'd be totally right to point that out. I just felt like getting it out of my head.
If most of the code you write is meaningful code that's novel and interesting then you are incredibly privileged. Majority of code I've seen in the industry is mostly boring and a lot of it just boilerplate.
Absolutely, coders should be spending time developing new and faster algorithms, things that AI cannot do, not figuring out the boilerplate of a dropbox menu on whatever framework. Heck, we dont even need frameworks with AI.
meaningful code that’s novel and interesting then you are incredibly privileged
This is possible but I doubt it. It's your usual CRUD web application with some business logic and some async workers.
So then you do write a bunch of boilerplate such as HTTP endpoints, database queries, and so on.
Not really. It's Django and Django Rest Framework so there really isn't a lot of boilerplate. That's all hidden behind the framework
It's more that the iterative slop generation is pretty energy intensive when you scale it up like this. Tons of tokens in memory, multiple iterations of producing slop, running tests to tell it's slop and starting it over again automatically. I'd love the time savings as well. I'm just saying we should keep in mind the waste aspect as it's bound to catch us up.
I don't really find the waste argument terribly convincing myself. The amount of waste depends on how many tries it needs to get the answer, and how much previous work it can reuse. The quality of output has already improved dramatically, and there's no reason to expect that this will not continue to get better over time. Meanwhile, there's every reason to expect that iterative loop will continue to be optimized as well.
In a broader sense, we waste power all the time on all kinds of things. Think of all the ads, crypto, or consumerism in general. There's nothing uniquely wasteful about LLMs, and at least they can be put towards producing something of value, unlike many things our society wastes energy on.