Amazon shows how automation actually works under capitalism - the robots do all the easiest jobs so humans are forced to do harder and harder work for less and less pay, and the robots break all the time and so the humans have to pick up their slack.
: “Whoa, what do you MEAN you expected to live a life off of being a worker! I like business so everyone should now all be their own little businessman and start their own businesses! I made labor free, you’re welcome!”
Calling it now, all the romanticization of being a one-man entrepreneur is going to go up to 11 pretty soon.
Hell, I’m seeing some of it already with all the ads marketing products “for your business” and all the shit job advice for people who can’t get one to simply start their own business.
They will never be able to fully automate it.
To contradict @queermunist@lemmy.ml abit here, what automation allows for is the simplification and breakdown of repetitious tasks, such as monitoring or sorting. Sometimes these can be complex in nature, but simplified through automation. However, atart up, shut down, and maintenance are still human tasks and will remain so for the near and far future, alongside design, most visual quality control checks, and detail work.
However, to agree with what she said part of the problem is that companies will often get contracts under the assumption that the automation works perfectly, which as an engineer, trust me it never does, and businesses always short-staff their maintenance and automation departments because they hate the idea of someone sitting around waiting for something bad to happen (inefficiency to them) even though that is the proper workflow for those areas, which is hurry up and wait. It’s wild that they understand that for sales and quality control, but if it’s a blue collar worker suddenly they have to be sweating their ass off 24/7 or they are a drain on the company.
True automation will require a complete redesign of factories and machines on a scale unfathomable under capitalism.
China has certainly come the closest. That said, maintenance, start-up and shut down are still human operations on most of those lines. However, for many of them they are at the point where they only need sensors and not visual confirmation, which is leagues ahead of where most American factories are.
It is actually much easier to automate most manufacturing, but the profit margins aren’t quite high enough to push it through. More money to be made on finance gambling and rent seeking.
Skipping over a great deal of explanations, we know from the labor theory of value that value can be added when making a product when socially necessary labor is being employed in its manufacture. If something can be completely automated from start to finish, then the value added will tend to zero rather quickly as the labor needed to make something of this type becomes zero. Crucially, not everyone needs to adopt automation for the social necessity of the labor to change. Only one or a few firms need to do it for this change to take place and then typically the others will be outcompeted over a short period as the profit they can take goes to shit.
This is why I find the trend of using AI to automate software engineering rather amusing as a software engineer. If completely successful it could remove one of the best paying jobs from the US economy entirely as the production of software by humans would no longer be socially necessary, the same as other things that can be automated. Then software engineers can all fulfil their destinies of becoming Factorio Youtubers.
Yes, that is how automation functions in competitive industrial capitalism. The key is that finance capital is in charge and has been for ages. They are already monopolies, or really owners of diversified monopolies, looking to steal each other’s lunches and scrape the barrel for profit via leveraging.
A simple example is Uber. Sure, having an app to get a taxi seems like an automation that bested the competition. But really it was the financial leveraging that led the way: it was cheap, the finance companies subsidized the price to drive market share, then raised prices once they had monopolies. This tech-finance combination is the heart of silicon valley, and is also federally subsidized. It is simultaneously key for US worldwide tech dominance.
“AI” is primarily driven in this way as well. It is massively overleveraged and the actual tech is not actually that useful, it doesn’t really make a particularly better product and the baseline to which it does it underwhelming. But they hope that there will be some general use case that solidifies it so that they can be one of the monopolies that jacks up prices once they have the market share. One of their major potential customers is C-suites at tech companies, people who are (1) hype beasts who desperately want to market their company as “with it” and therefore purchasable for a high price by the monopolies, (2) themselves over leveraged on finance and desperate to smugly fire 75% of their workforce if their finance papers can still look good at the end, and (3) are extremely gullible when it comes to any topic with these qualities. The key here is that “AI” doesn’t even need to competently automate, and with software, it truly does not. It’s a faster prototyper, maybe, but it makes so many fundamental errors, of a predictable-yet-not-fixable-by-a-non-expert type, that you can’t actually fire any significant number of people and get the same productivity. It really just clarifies the role of engineers, which is not code monkeying, but identifying project scope, planning, maintaining, security, and sometimes, rarely, solving algorithmic or math problems. An example of this is C-suites showing everyone the “app” they vibe coded that is basically just a more verbose tutorial app with some menus. Congratulations Brad, you really did it. You made an app that doesn’t do any of the things spec’s out by your product team as core necessities because you were spending half your time for the last month not thinking about that at all and instead begging a prompt to write code. The core issue of clarity of thought and planning, or really the absence of it, is laid bare. And God help you if data analysis is required.
Very cool, looking forward to when the manufacturing is also automated.
Amazon shows how automation actually works under capitalism - the robots do all the easiest jobs so humans are forced to do harder and harder work for less and less pay, and the robots break all the time and so the humans have to pick up their slack.
In about 10 years later:
Calling it now, all the romanticization of being a one-man entrepreneur is going to go up to 11 pretty soon.
Hell, I’m seeing some of it already with all the ads marketing products “for your business” and all the shit job advice for people who can’t get one to simply start their own business.
And you’ll “own your own business” the same way that you’re currently an “independent contractor” working for Uber.
They will never be able to fully automate it. To contradict @queermunist@lemmy.ml abit here, what automation allows for is the simplification and breakdown of repetitious tasks, such as monitoring or sorting. Sometimes these can be complex in nature, but simplified through automation. However, atart up, shut down, and maintenance are still human tasks and will remain so for the near and far future, alongside design, most visual quality control checks, and detail work.
However, to agree with what she said part of the problem is that companies will often get contracts under the assumption that the automation works perfectly, which as an engineer, trust me it never does, and businesses always short-staff their maintenance and automation departments because they hate the idea of someone sitting around waiting for something bad to happen (inefficiency to them) even though that is the proper workflow for those areas, which is hurry up and wait. It’s wild that they understand that for sales and quality control, but if it’s a blue collar worker suddenly they have to be sweating their ass off 24/7 or they are a drain on the company.
True automation will require a complete redesign of factories and machines on a scale unfathomable under capitalism.
isnt this going on in china rn with all of those “dark” factories?
China has certainly come the closest. That said, maintenance, start-up and shut down are still human operations on most of those lines. However, for many of them they are at the point where they only need sensors and not visual confirmation, which is leagues ahead of where most American factories are.
It is actually much easier to automate most manufacturing, but the profit margins aren’t quite high enough to push it through. More money to be made on finance gambling and rent seeking.
Skipping over a great deal of explanations, we know from the labor theory of value that value can be added when making a product when socially necessary labor is being employed in its manufacture. If something can be completely automated from start to finish, then the value added will tend to zero rather quickly as the labor needed to make something of this type becomes zero. Crucially, not everyone needs to adopt automation for the social necessity of the labor to change. Only one or a few firms need to do it for this change to take place and then typically the others will be outcompeted over a short period as the profit they can take goes to shit.
This is why I find the trend of using AI to automate software engineering rather amusing as a software engineer. If completely successful it could remove one of the best paying jobs from the US economy entirely as the production of software by humans would no longer be socially necessary, the same as other things that can be automated. Then software engineers can all fulfil their destinies of becoming Factorio Youtubers.
Yes, that is how automation functions in competitive industrial capitalism. The key is that finance capital is in charge and has been for ages. They are already monopolies, or really owners of diversified monopolies, looking to steal each other’s lunches and scrape the barrel for profit via leveraging.
A simple example is Uber. Sure, having an app to get a taxi seems like an automation that bested the competition. But really it was the financial leveraging that led the way: it was cheap, the finance companies subsidized the price to drive market share, then raised prices once they had monopolies. This tech-finance combination is the heart of silicon valley, and is also federally subsidized. It is simultaneously key for US worldwide tech dominance.
“AI” is primarily driven in this way as well. It is massively overleveraged and the actual tech is not actually that useful, it doesn’t really make a particularly better product and the baseline to which it does it underwhelming. But they hope that there will be some general use case that solidifies it so that they can be one of the monopolies that jacks up prices once they have the market share. One of their major potential customers is C-suites at tech companies, people who are (1) hype beasts who desperately want to market their company as “with it” and therefore purchasable for a high price by the monopolies, (2) themselves over leveraged on finance and desperate to smugly fire 75% of their workforce if their finance papers can still look good at the end, and (3) are extremely gullible when it comes to any topic with these qualities. The key here is that “AI” doesn’t even need to competently automate, and with software, it truly does not. It’s a faster prototyper, maybe, but it makes so many fundamental errors, of a predictable-yet-not-fixable-by-a-non-expert type, that you can’t actually fire any significant number of people and get the same productivity. It really just clarifies the role of engineers, which is not code monkeying, but identifying project scope, planning, maintaining, security, and sometimes, rarely, solving algorithmic or math problems. An example of this is C-suites showing everyone the “app” they vibe coded that is basically just a more verbose tutorial app with some menus. Congratulations Brad, you really did it. You made an app that doesn’t do any of the things spec’s out by your product team as core necessities because you were spending half your time for the last month not thinking about that at all and instead begging a prompt to write code. The core issue of clarity of thought and planning, or really the absence of it, is laid bare. And God help you if data analysis is required.
I’m getting a message on my walkie talkie. It sounds like it has already been done. Well, nevertheless.
deleted by creator