Ok, but you should hate it though. AI, as implemented by capitalism, is downright detrimental in so so so many ways. Even if we set aside the huge environmental costs, energy and water price hikes in communities near data centers, loss of jobs due to AI, theft of IP, and sloppification of the internet, it is also doing things like convincing my friend with bipolar that they should attempt suicide.
That’s kinda contained in “the CEOs forcing it down our throats” though, y’know? They’re forcing it down our throats because capitalism requires it, they need public buy-in and mass adoption or else the whole economic structure implodes.
None of that tells me I should hate the technology, like any other technology under capitalism.
You can’t just take things out of their historical context, or you’re not doing dialectics anymore. We live in the reality we live in. However, even if we lived in a communist utopia, the tool’s insane resource requirements and impacts on people’s health and livelihoods should still make you disdain its use in any context other than research.
We also live in a historical moment where there are national experiments with different material relations than the predominant capitalist form of the West. China specifically is doing things somewhat differently by training models to be modular and task-specific, instead of the obsession Silicon Valley has with making an everything app. I’m also interested in their experiments in cooling data centers with sea water, as well as how they’re phasing out fossil fuels while rapidly expanding next gen renewables+grid battery storage to power those data centers.
I don’t know if I’ll go as far as to say they’re doing AI right, but I think when the bubble bursts they’ll adapt far better.
Good point, and I don’t really know how LLMs are being rolled out in China. However, given the context of climate change and the impending broader global ecological collapse, I really don’t see any reason at all that a tool with such limited usefulness relative to its outsized footprint should be made broadly available. It’s like, on a global scale we are already doing basically nothing about how completely fucked we’re going to be in the 30-70 years.
Ok, but you should hate it though. AI, as implemented by capitalism, is downright detrimental in so so so many ways. Even if we set aside the huge environmental costs, energy and water price hikes in communities near data centers, loss of jobs due to AI, theft of IP, and sloppification of the internet, it is also doing things like convincing my friend with bipolar that they should attempt suicide.
That’s kinda contained in “the CEOs forcing it down our throats” though, y’know? They’re forcing it down our throats because capitalism requires it, they need public buy-in and mass adoption or else the whole economic structure implodes.
None of that tells me I should hate the technology, like any other technology under capitalism.
You can’t just take things out of their historical context, or you’re not doing dialectics anymore. We live in the reality we live in. However, even if we lived in a communist utopia, the tool’s insane resource requirements and impacts on people’s health and livelihoods should still make you disdain its use in any context other than research.
We also live in a historical moment where there are national experiments with different material relations than the predominant capitalist form of the West. China specifically is doing things somewhat differently by training models to be modular and task-specific, instead of the obsession Silicon Valley has with making an everything app. I’m also interested in their experiments in cooling data centers with sea water, as well as how they’re phasing out fossil fuels while rapidly expanding next gen renewables+grid battery storage to power those data centers.
I don’t know if I’ll go as far as to say they’re doing AI right, but I think when the bubble bursts they’ll adapt far better.
Good point, and I don’t really know how LLMs are being rolled out in China. However, given the context of climate change and the impending broader global ecological collapse, I really don’t see any reason at all that a tool with such limited usefulness relative to its outsized footprint should be made broadly available. It’s like, on a global scale we are already doing basically nothing about how completely fucked we’re going to be in the 30-70 years.