TL;DR: AI keeps getting massive investment despite not working as promised.
-
RE: https://kolektiva.social/@danmcquillan/116098694142499511
TL;DR: AI keeps getting massive investment despite not working as promised. This is because AI isn't really a technology solution, but instead a way for existing power structures to maintain control as the old world order falls apart. It makes the problems it claims to solve *worse* rather than better.
The alternative is "decomputing": less tech, instead of more; slower, not faster; tools built around collective care rather than automation.
Instead of asking how to power bigger data centres, we should ask who should control society's energy and water in the first place.
-
RE: https://kolektiva.social/@danmcquillan/116098694142499511
TL;DR: AI keeps getting massive investment despite not working as promised. This is because AI isn't really a technology solution, but instead a way for existing power structures to maintain control as the old world order falls apart. It makes the problems it claims to solve *worse* rather than better.
The alternative is "decomputing": less tech, instead of more; slower, not faster; tools built around collective care rather than automation.
Instead of asking how to power bigger data centres, we should ask who should control society's energy and water in the first place.
To which my response is, yes, but that questioning isn't in the hands of individual users but rather on the level of nation states.
So calling on people to 'decompute' is a rallying cry to the kind of opting out that might make sense in terms of appealing to people's ethics, but is interminably difficult under capitalism.
-
To which my response is, yes, but that questioning isn't in the hands of individual users but rather on the level of nation states.
So calling on people to 'decompute' is a rallying cry to the kind of opting out that might make sense in terms of appealing to people's ethics, but is interminably difficult under capitalism.
(I didn't simplify the verbose academese to agree with it)
-
R relay@relay.infosec.exchange shared this topic
