I have to start with some pushback: I know you don’t mean it that way, but these types of comments are dismissive of all the work comrades put into using and developing AI models to solve actual problems, such as making more theory/marxist writing available (especially pertaining to local conditions). Every time a comrade talks about how they use AI to solve problems for the movement, there’s always one person that chimes in trying to rain on their parade, and I told myself some time ago that I would start calling it out when I see it.
The environmental impact is no worse than anything else you consume and use under capitalism already. It’s difficult to properly gauge the environmental impact of anything, especially when the service is digital and new like AI is. But there are much more pressing industries to tackle if we want to do something about the environment, and as communists we know that individual responsibility is a myth invented by fossil fuel companies. As soon as the word AI is pronounced some people retreat to liberal mythology about how there’s suddenly ethical consumption, change at the local level leads to change globally, and if people just had the right ideas and habits we would fix the world.
Us refusing to use AI for our own needs will not suddenly make it disappear. I wrote an essay about this; it will continue doing what it does regardless of how much we posture against it, because posturing does not produce material change. Should the bourgeoisie have the monopoly on this new technology, or should we get in there and develop and use it for our own goals?
As for translation work, I have used it in languages I speak and it works well enough. Otherwise I wouldn’t recommend it obviously. It’s not going to be as good as a professional translator would get, but:
a professional translator at the level that it sounds native will take weeks to translate just an essay. AI takes a couple minutes of waiting.
we are not all professional translators. Even speaking the language natively I don’t think I could do a better job than the AI.
hiring one is expensive. The cost is prohibitive for most people and only makes sense commercially.
it frees up time to do other things.
more importantly, it’s the difference between having something produced and not having anything.
Prior to LLMs I tried for 3 years to crowdsource a book translation. Our grand total of pages translated was 0 for a team of 12 at its peak. One night two of us took it to an old version of chatGPT and got it done in 2 hours.
This is a force multiplier for a movement who is always short on time and people.
But AI is really bad for the environment, and is it even that good for translation?
I have to start with some pushback: I know you don’t mean it that way, but these types of comments are dismissive of all the work comrades put into using and developing AI models to solve actual problems, such as making more theory/marxist writing available (especially pertaining to local conditions). Every time a comrade talks about how they use AI to solve problems for the movement, there’s always one person that chimes in trying to rain on their parade, and I told myself some time ago that I would start calling it out when I see it.
The environmental impact is no worse than anything else you consume and use under capitalism already. It’s difficult to properly gauge the environmental impact of anything, especially when the service is digital and new like AI is. But there are much more pressing industries to tackle if we want to do something about the environment, and as communists we know that individual responsibility is a myth invented by fossil fuel companies. As soon as the word AI is pronounced some people retreat to liberal mythology about how there’s suddenly ethical consumption, change at the local level leads to change globally, and if people just had the right ideas and habits we would fix the world.
Us refusing to use AI for our own needs will not suddenly make it disappear. I wrote an essay about this; it will continue doing what it does regardless of how much we posture against it, because posturing does not produce material change. Should the bourgeoisie have the monopoly on this new technology, or should we get in there and develop and use it for our own goals?
As for translation work, I have used it in languages I speak and it works well enough. Otherwise I wouldn’t recommend it obviously. It’s not going to be as good as a professional translator would get, but:
Prior to LLMs I tried for 3 years to crowdsource a book translation. Our grand total of pages translated was 0 for a team of 12 at its peak. One night two of us took it to an old version of chatGPT and got it done in 2 hours.
This is a force multiplier for a movement who is always short on time and people.