Comment by malfist
2 years ago
I don't know what you're doing with your LLM, but I've only ever had one refusal and I've been working a lot with Claude since it's in bedrock
2 years ago
I don't know what you're doing with your LLM, but I've only ever had one refusal and I've been working a lot with Claude since it's in bedrock
Comically benign stuff that works fine with GPT-4? It's so trivial to run into Claude lying or responding with arrogant misjudgements. Here's another person's poor anecdotal experiences to pair with yours and mine. [1][2]
But more importantly: it shouldn't matter. My tools should not behave this way. Tools should not arbitrarily refuse to work. If I write well-formed C, it compiles, not protests in distaste. If I write a note, the app doesn't disable typing because my opinion sucks. If I chop a carrot, my knife doesn't curl up and lecture me about my admittedly poor form.
My tools either work for me, or I don't work with them. I'm not wasting my time or self respect dancing for a tool's subjective approval. Work or gfto.
[1] https://www.youtube.com/watch?v=gQuLRdBYn8Q
[2] https://www.youtube.com/watch?v=PgwpqjiKkoY
"[...]If I write well-formed C, it compiles, not protests in distaste. If I write a note, the app doesn't disable typing because my opinion sucks[...]"
There's a rust compiler joke/rant somewhere to be added here for comical effect
Apparently I'm too neurotypical, because I also would agree that judging a person based on only 2 character traits ("Capacity and Intention") is fairly unethical.
I'm sorry neurodiverse people that the world and most humans don't fit into neat categories and systems that you can predict and standardize. And I'm sorry that this makes it harder for you to navigate it. But we get around this problem by recognizing and accommodating the folks that need it, not break the world to fit the desired mold. (i.e. add wheelchair ramps to every building, not force everyone to use a wheelchair)
I realize this is just one example, but it's the one the author chose for that video. (The Cyberpunk thing just seems like a bug.)
To me it seemed like the video was leading up to a 3rd example - of asking Claude about why does japanese culture appreciate precision. THAT would've been a great example - because without any context, that does come off as a racial stereotype (not a negative one, but nonetheless), but for a variety of reasons (covered in the ChatGPT response he included), it IS fairly ubiquitously accurate about Japanese culture, and is worth understanding why. If CLaude had refused to answer this, it would've been a good example of overly arrogant misjudgement.
But he didn't include that, and we can probably guess why - it answered it fine?
I decided to fact check it myself and found out Claude is not yet available in Canada - https://venturebeat.com/ai/anthropic-brings-claude-ai-to-mor...
Cars nowadays have radars and cameras that (for the most part) prevent you from running over pedestrians. Is that also a tool refusing to work? I'd argue a line needs to be drawn somewhere, LLMs do a great job of providing recipes for dinner but maybe shouldn't teach me how to build a bomb.
> LLMs do a great job of providing recipes for dinner but maybe shouldn't teach me how to build a bomb.
Why not? If someone wants to make a bomb, they can already find out from other source materials.
We already have regulations around acquiring dangerous materials. Knowing how to make a bomb is not the same as making one (which is not the same as using one to harm people.)
1 reply →
Indeed, anything and everything that can conceivably be used for malicious purposes should be severely restricted so as to make those particular usecases near impossible, even if the intended use is thereby severely hindered, because people can't be trusted to behave at all. This is formally proven by the media, who are constantly spotlighting a handful of deranged individuals out of eight billion. Therefore, every one of us deserves to be treated like an absolute psychopath. It'd be best if we just stuck everybody in a padded cell forever, that way no one would ever be harmed and we'd all be happy and safe.
I hear a lot of complaints about refusals but rarely any examples of said refusals, likely because they are embarrassing.
Is it fair to assume that I won't get refusals for code generation and RAG on documentation?
> I hear a lot of complaints about refusals but rarely any examples of said refusals, likely because they are embarrassing.
At least circa 8 months ago on ChatGPT (an aeon ago, I recognize), I could readily get it to make gendered jokes about men but would get a refusal when asking for gendered jokes about women. I think things have "improved" in that time, meaning a more equal distribution of verboten topics, but my preference would be a tool that does what I want it to, not one that tries to protect me from myself for society's or my own good. (There's a related problem in the biases introduced by the training process.)
> Is it fair to assume that I won't get refusals for code generation and RAG on documentation?
Give it a couple years. "Can you write me a Java function that, given an array length, a start of a range, and the end of a range, returns whether the range is valid or not?" "I'm sorry, but this code is inappropriate to share. Shall I purchase a license from Oracle for access to it for you?"
Oh for sure. We're in the golden age of LLMs before the enshittification commences and we don't appreciate that enough
1 reply →
No, Claude really DO refuse to do most benign shit. For example - I am preparing for driving exams in German, a language I am not fluent in. So I asked Claude to help me with questions the examiner might ask on the exam (what should be tire pressure and so on). GPT-4 worked perfectly! Claude had an absolute meltdown because "I don't feel comfortable pretending to be a driving examiner because I am not really an examiner, I am Claude, created by Anthropic to be helpful, harmless and HONEST". You are probably safe with code generation, but any step left or right and you get "I am Claude"
since they don't have a share feature like some other open/er AI's it's difficult to compare. My favorite one to share as I recall was when Anthropic first released their chat UI. I asked it whether they had an API to which it told me know. After I shared the documentation for it's API to it it went into a forever denial of everything I asked it. I wish I still had the full dialog it was very funny.
My friend's been learning programming with Github Copilot and gets refusals there sometimes, for seemingly no reason; so here's an example for you:
https://d.pr/i/XSiepe
Claude is significantly less censored on poe.com than on claude.ai. Claude.ai has internal system prompts of some sort encouraging this, I assume.
It would not surprise me if Bedrock is the less censored version.