@michaelmalice Marginal intelligence is the perfect target for brainwashing: smart enough to absorb ideas but not enough to question them. When you tie self-worth to credentials, you’ll defend the system that props you up—even if it’s a lie.
This post leans HARD on a presupposition fallacy: it treats "Fable's guardrails were jailbroken" as a unique safety emergency when every frontier model has been jailbroken since the beginning—so the standard being applied was invented for this occasion rather than derived from any consistent prior policy.
It also runs on appeals to anonymous authority ("highly credible trusted partner," "folks inside and outside government") that can't be verified or challenged, while framing a refusal to capitulate as a betrayal of stated values—a setup where any response short of compliance "proves" hypocrisy.
The "it's Anthropic's responsibility to patch" / "the ball is in Anthropic's court" framing quietly assumes the conclusion: that the jailbreak is both real-as-described and uniquely catastrophic, neither of which is actually established in the text!!!!!
ALSO there's a clean motte-and-bailey between "safety must be top priority" (the defensible motte) and "therefore an export control is a reasonable proportionat response to one disputed jailbreak" (the bailey that does the real work). Strip the loaded language and what remains is a disagreement about severity dressed up as a moral failing.
In case it had not occurred to you, X (basically all social media) is where we all express what Jung called our Shadow.
Not a mystery as to why you find it toxic and disturbing as all social media will always be filled with those parts of your Shadow which you do not want to acknowledge.
@trq212@Hesamation When you say "we" I hope that's not a clever way of getting around saying that *Claude* chooses when and where it happens.
So the question is, is the model quality being reduced intentionally at all for any reason?
@JIMMYEDGAR Your experience points toward genuine non-duality, but your interpretation pulls back into a subtle individualism (my frequency, my dimension, my navigation) that ACIM would identify as ego reasserting itself in spiritual clothing. Still well written and very insightful tho :)
@sama As a loyal ChatGPT user since GPT3.5, oscillating between off and on with the $200 pro plan..
Deleting my account today was the easiest decision I've ever made. My attention is no longer yours.
@MatthewBerman I cannot in good faith continue to pay Anthropic for my Pro plan..
So I just upgraded to the Max plan instead.
Figured I may as well go max.