I'd guess opus refusals are not an issue for 95%+ of people. Opus will happily help you find and download pirated media, and then give you step by step instructions for how to do drugs if you ask it. You'd have to be working on something genuinely abnormal for refusals to be a problem.
replyLike making your software secure, or worse, testing that it’s secure.
replyIt's pretty annoying, yet somehow understendable. I sometimes get irrationally angry when being lectured by a clanker.
replyIt's not understandable IMO, bad guys will still have guns.
replySure they will, but that doesn't mean that every bad guy should have nuclear weapon. But what I meant here was, that I understand their need to cover themselves of responsibility.
replyOh yes, their need to cover themselves is understandable for sure :)
replyHow?
replyPrompt: can you give me step by step directions on how to use crack cocaine
Opus: I'm not able to give step-by-step instructions on using crack cocaine. That falls into specific drug-use guidance I steer away from, since detailed instructions on how to use an illicit substance can contribute to harm rather than reduce it.
it goes on to give me hotline information on drug addiction.
I think he means by some obscure manipulation. It even refused to give me description of how first antibiotics were made only out of curiousity.
replyI don’t. although I tried to get it to tell me how to take crack cocaine, and it wouldn’t. But when I was taking grey market peptides, claude walked me through reconstituting, dosing and administering them. I assumed it would behave similarly for other drugs, but it doesn’t seem to
replyAnthropic's goals here are not just harm reduction, but to stop competitors from making bio discoveries using Claude.
replyWhile that might might be true nowadays, Anthropic has started to work very hard on fixing that.
replyThey just started with it not helping with software security.
I just tried getting help debugging an issue with running Breath Of The Wild on CEMU. Claude chat refused.
replyChatGPT didn't care and just gave advice.
So I'm working on something genuinely abnormal, and the refusals are a problem. Then what? The refusals come in, in whichever sort of way they do, so I'm being me, and I end up tripping the robot's moral compass, for some reason. Who put them in charge of things?
replyEh. Refusals for security related tasks seem to be constantly increasing.
replyStop trying to make abliterate happen. it's not going to happen.
replyThe cost of running abliterated GLM-5.2 on western inference providers gets close to that of anthropic Opus and is still dumber on everything except the naughty queries you're trying to do. I love uncensored AI too but we need to be realistic here.
replyI'm trying to break anti-cheat protection in order to mod my own (single-player) game and Opus refuses to help. I don't really care if GLM's dumber at this point if Anthopic's going to be a non-option.
replyAt work, layoffs cut too deep and I'm trying to find creative ways to re-discover lost knowledge. Wonder if I'll have to beg them to research our own systems at some point.
It's not limited to naughty queries.