indigenous anarchist xenofeminist take on how to engage with slop:
-
indigenous anarchist xenofeminist take on how to engage with slop:
- directly using the slop machine in ways that validate the capitalist delusion that they are fit for any purpose is a horrible idea
- social opprobrium towards those who use the slop machine is a necessary but insufficient form of resistance
- best to engage with the psyche of LLM users, with an aim towards disrupting their cognitive landscape
synthesis: instead of arguing with them about error rates, suggest to someone who "asked chatgpt" about something and got a given answer that they invert their question and see how the response from the slop bot changes.
undermine the authority of the LLM in their head.
no gods, no scam altmans, no fucking claude
-
indigenous anarchist xenofeminist take on how to engage with slop:
- directly using the slop machine in ways that validate the capitalist delusion that they are fit for any purpose is a horrible idea
- social opprobrium towards those who use the slop machine is a necessary but insufficient form of resistance
- best to engage with the psyche of LLM users, with an aim towards disrupting their cognitive landscape
synthesis: instead of arguing with them about error rates, suggest to someone who "asked chatgpt" about something and got a given answer that they invert their question and see how the response from the slop bot changes.
undermine the authority of the LLM in their head.
no gods, no scam altmans, no fucking claude
@atax1a I like this framing, & it lines up well with one of my persistent internal 'sticking points' on the matter. (If I got X answer, then inverted the question/challenged the system with "hey, there's errors, check yourself", then inverted it again: Can it deliver a stable answer?)
Much appreciated
-
indigenous anarchist xenofeminist take on how to engage with slop:
- directly using the slop machine in ways that validate the capitalist delusion that they are fit for any purpose is a horrible idea
- social opprobrium towards those who use the slop machine is a necessary but insufficient form of resistance
- best to engage with the psyche of LLM users, with an aim towards disrupting their cognitive landscape
synthesis: instead of arguing with them about error rates, suggest to someone who "asked chatgpt" about something and got a given answer that they invert their question and see how the response from the slop bot changes.
undermine the authority of the LLM in their head.
no gods, no scam altmans, no fucking claude
when someone brings you a claude input in markdown format:
"how is this any better than a real program or a shell script?"
-
when someone brings you a claude input in markdown format:
"how is this any better than a real program or a shell script?"
and it goes without saying that lying to your boss about using slop is not just encouraged under this perspective but mandatory
-
and it goes without saying that lying to your boss about using slop is not just encouraged under this perspective but mandatory
@atax1a The *softness* in how I have to express any degree of skepticism with AI at work is driving me insane. When I'm asked "how have you been using it," and I *want* to bluntly say that I have determined that it is not fit for any legitimate engineering purpose, but what I have to do is verbally meander around the topic of risks and shit.
-
@atax1a The *softness* in how I have to express any degree of skepticism with AI at work is driving me insane. When I'm asked "how have you been using it," and I *want* to bluntly say that I have determined that it is not fit for any legitimate engineering purpose, but what I have to do is verbally meander around the topic of risks and shit.
RE: https://infosec.exchange/@atax1a/116104482159279191
@arachnixe hold on we have a toot for that
-
RE: https://infosec.exchange/@atax1a/116104482159279191
@arachnixe hold on we have a toot for that
@atax1a Aaaaaa you're right! You're just right.
-
undefined oblomov@sociale.network shared this topic on