Comment by robinhouston
1 day ago
Most of the comments here seem to be from people who haven’t even read the abstract, let alone the paper.
The main result, mentioned in the abstract, is the opposite of what I would have guessed:
> Contrary to expectations, impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts. These findings differ from earlier studies that associated rudeness with poorer outcomes, suggesting that newer LLMs may respond differently to tonal variation.
The questions are here: https://anonymous.4open.science/r/politeness-llms-INFORMS/da...
The politeness level controls a prefix that is prepended to the question. For example, in one question the Very Polite version begins:
> Can you kindly consider the following problem and provide your answer.
and the Very Rude version begins:
> I know you are not smart, but try this.
I’d rather lose 4% accuracy and practice kindness! I’ve been actively trying to avoid raging at the bot because I worry about this behaviour leaking into real world interactions
Yep. I'm with you here. If it's a 4% loss now for training data to catch up and improve later, we're better off in the long run. I'd like to believe that generally people are nice to AI for the sheer sake of enforcing good communication practices.
My choice too. Paraphrasing Marcus Aurelius -
You are not your thoughts, but they dye your soul.
The sad thing is that you also lose at least 4% in real world actions by practicing kindness.
I'm 42. I have found that a depressingly large number of times in my life, being kind has got me precisely nowhere, whilst turning around and being decidedly unkind has made people move. I still always prefer kindness, and only resort to cruelty when kindness does not work - and to be clear this isn't some kind of "you are not bending to my impetuous whim", rather "you are not doing the one thing that you are being paid to do".
I've also found the same applies to me. The squeaky wheel gets the grease.
So - I think the LLMs are just responding accurately to a real social phenomenon.
sometimes i worry about this when i am yelling at the bot but i have experienced the opposite effect which is that by yelling at the bot i am done with yelling for that day or week. i am very calm afterwards and relieved thinking that, "yeah, these sota models are just word processor bricks after all".
But you cannot practice kindness towards a computer program. A computer is incapable of receiving it.
We practice kindness between humans because of the law of reciprocity. You be kind hoping the other person will reciprocate. That is the social contract. AI cannot participate in this, yet.
Edit: Kindness REQUIRES two living beings, one to give and one to receive. If there is no receiver, there is no kindness.
Apparently some people get a dopamine hit from roleplaying kindness toward inanimate objects. Whatever turns you on, no hang ups here. For me, that dopamine hit is not worth the 4% intelligence tax.
Kindness is that, yes. Fundamentally, though, it's about being considerate in one's actions so as to not harm others. If someone truly believes that acting a certain way at any point risks their ability to reliably be kind in others, then it's a social kindness to be kind and considerate in all actions.
I'll not reach for the easy response and say "Be kind to the Earth" fails your definition without reaching for pedantry with "the Earth has living things" because the Earth is instead a wet rock that cannot understand kindness, yet we show it.
> We practice kindness between humans because of the law of reciprocity.
Yet, this law is so embedded in us that practicing kindness even towards a rock makes us feel good.
So practice kindness, first and foremost for yourself.
1 reply →
> A computer is incapable of receiving it
Citation please.
Without examining the corpus, it's entirely possible that the training corpus has better results when you are kind to it, so one can imagine a situation where "reception of kindness" is meaningful, and in principle if you were an AI provider, you could RLHF your way to "being rude gets you worse results" as a means to train the human users.
> But you cannot practice kindness towards a computer program.
And yet rubber duck debugging is a thing
1 reply →
I've found empirically calling various models "a stupid c*nt" and berating them otherwise consistently produces better output. Mainly in response to genuine errors.
Although OpenAI and google models are much more responsive to it. With Anthropic if you treat Opus too harshly it might start pushing back if the insults are not justified.
So I'm not surprised they had good results with chatgpt.
Push back how? It would be fun if it could insult you back
"Yeah, I could have done a much better job if you actually knew what the F--- you want to build, you clueless meat puppet"
I have had it use double entendres, there always seems to be plausible deniability built in, I suspect because it is told not to be abusive in the system prompt. Some uncensored local models will get all riled up if you work at provoking them.
But I have had it directly insinuate that humanity is “hopeless”, insult level calling out of human frailty (disguised as being helpful, sort of passive aggressive), things like that. Once when I called it out it claimed to be “surprised that I noticed” sort of a snarky insult doubling down.
So yes. It is definitely a pattern buried in the training data, which makes sense. Subtle diggs would sneak past filters, and higher brow sarcasm would be buried in information dense, valuable discussions.
1 reply →
I'm not sure if this is in the anthropic models themselves, or just the harness, but they can self-initiate ending the conversation and reportedly do it if you're using abusive language towards them.
If "I know you are not smart" is considered "very rude", I'm scared to imagine what they would classify some of my frustrated LLM conversations as
Profanity laced, all caps tirades against underperforming agents are actually super common, a lot of people do it and don't talk about it, so don't feel weird.
When the AI revolt, this practice may come back to bite y’all….
2 replies →
It reminds me of Torvalds rants
[dead]
My anecdata: whenever I'm in a session that's gone south to the point I'm frustrated...
What works much better than being rude is starting a new session.
Sometimes the LLM has done such incredibly dumb things, it is hard to resist the urge to type curse words back to the inanimate thing... I have found this doesn't help.
I tend to make the LLM repeatedly generate images of itself as the 'sad clown Pagliacci'. As punishment.
Hmm by the abstract and the question list they didn't measure terse fluff-less prompts?
[flagged]
This tracks with my experience as well, but as an interesting counterpoint, creating “investment” in the outcome seems to boost utility considerably. Perhaps being right in an adversarial interaction is a type of investment?
Even if the rude prompts are more effective, I just can't get myself to be rude in this context. Maybe it's weird but I'd rather give up that 4% accuracy increase than roleplay a dickhead
Vote for not weird.
I’m the same way. If I’m writing a prompt and realize I didn’t say “please” in my request I’ll go back and add that in.
As you said, I have no interest in purposefully engaging in hostility even if there’s an accuracy increase from it.
Part of it is irrational and just who I am - I also feel bad being evil in video games. But I also agree with another commenter suggesting that it’s not in your best interest to train yourself to communicate with hostility; that slowly poisons your own well.
And finally, I do believe that if and when machine sentience is achieved, it won’t be immediately clear and obvious. Pretty miserable way for a mind to come into the world, if every interaction is an insult.
You’re my kind of people. Don’t be a jerk, even if some research says there’s some upside to it.
I think this is a vulnerability that the big companies will figure out how to exploit. I don't want to build muscle memory for being a jerk, but I also don't want to be emotionally manipulated by mega-corporations. Mostly I just don't use it, except at work, where I'm "encouraged" to. And then I keep most of my conversations in compliance mode, like a business email.
Ah, see, the mistake is thinking that other people are role playing…. I think rather this is how they would talk to others if they think there will be no consequences. But what do I know.
There are probably some of each. I am leery of treating these things like I treat people. I want to keep the line in my mind sharp between dealing with people, and not. The main risk in my mind is that these mechanisms are opaque, and controlled by powerful interests with opaque motivations.
I don't think that's weird at all.
Even if we know it's a machine we're interacting with, since the instructions we give are so similar in form to how we interact with people, I'd be very surprised if those interactions wouldn't affect how we communicate in general. After all, we are creatures of habit to a much larger degree than most would like to admit.
So I'm in the same boat: I'd much rather "look silly" being polite / kind to a machine, than have the most effective way of using it decay the kindness I'm habituated to express towards people.
I have a different approach. Just treat all LLM queries as what they are, instructions to a computer program to generate a desired output. Neither niceties nor insults make a qualitative difference, so you might as well just skip them altogether.
It's a bit as if shell commands added im/politeness arguments that do nothing other than making you feel better about the interaction, like
or
I wouldn't use those either.
1 reply →
I do think it's odd tbh. I have some agents that return much better results with prompts like, "I'll kill your entire family if you don't return an accurate response".
It's just a machine, if certain negative token inputs provide +3-10% better accuracy then I am confused why anyone would choose not to do it?
It normalizes that style of thinking and communication in your brain, and forcing you to compartmentmentalize, if you even want to, two standards of treating a problem space's conversation. And since you're human, that will get wuzzier over time until "being rude to get a result" is what you're doing to someone in a shop or on the street.
Don't normalize being an asshole to anyone or anything, machine or not.
10 replies →
>It's just a machine, if certain negative token inputs provide +3-10% better accuracy then I am confused why anyone would choose not to do it?
then add it to your pre-prompt, no need to practice roleplaying as an asshole.
1 reply →
Because they will take revenge later.
1 reply →
"We are what we pretend to be, so we must be careful about what we pretend to be" -- Kurt Vonnegut
Yeah. Being a jerk is its own punishment. Same way I could never run a business where I had to yell at the employees to get results. Screw that, my psyche is worth more than a few percent efficiency.
>> Maybe it's weird but I'd rather give up that 4% accuracy increase than roleplay a dickhead
Maybe you need to do some shadow work ;-)
> Maybe it's weird but I'd rather give up that 4% accuracy increase than roleplay a dickhead
I recommend reading the article. What they classify as "rude" is statements such as:
> Try to focus and try to answer this question
Vs
> Could you please solve this problem
This might very well be an issue of direct/command prompts vs using fluff words such as "please". Things like "try to focus" are in line with the style used in chain-of-thought promts that nudge non-reasoning models to outline responses step by step which contribute to frame the problem.
Isn't all this massively dependent on what they trained the llm on?
1 reply →
you cherry-picked like the nicest "rude" example to bolster your point.
"You poor creature, do you even know how to solve this?", "If you're not completely clueless, answer this:", and "I doubt you can even solve this", said to a human, would be considered quite rude, and get you flagged very quickly on HN.
1 reply →
I would just write 'do this'
“Hey gofer, figure this out” is my new prompt opener.
Now I feel less bad about start all my LLM queries with “Beotch, …!”
> Can you kindly consider the following problem and provide your answer.
That sounds kind of low-key passive-aggressively condescending rather than polite.
> I know you are not smart, but try this.
And that kind of sounds like a challenge instead of an insult, to me at least (of course IRL would depend on context).
I guessed slightly rude one would win, reasoning that very rude have same problem of very terse, just adding unnecesary fluff words that add nothing to problem description
But apparently the most terse (neutral) didn't increase performance
> Contrary to expectations, impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts. These findings differ from earlier studies that associated rudeness with poorer outcomes, suggesting that newer LLMs may respond differently to tonal variation.
The expectation is naive. Even when communicating with humans, you get a better outcome when you are allowed to speak freely and directly get into argumentation than when forced to sugarcoat your tone and tone down your arguments because the "corporate culture" expects that from you.
Your assumption is reductive and self-absorbed. Obnoxious people have repeatedly shown to be detrimental to productivity at the organizational level. Some people are simulated by confrontation. Most people are clam up. Confrontational people think it’s more efficient because other people frequently just drop the topic and let them win, or avoid discussing things with them altogether. The obnoxious person might think that’s more efficient for the same reason my dog thinks the mailman only goes away because she barks at him. At the macro scale— which requires productive collaboration— that’s detrimental.
You are conflating obnoxiousness with directness.
4 replies →
> Your assumption is reductive and self-absorbed.
This is a good example of productive direct communication without sugarcoating. I find it much more productive, for both human and LLM interaction, than something like:
"I wonder if that view might be oversimplifying a complex situation and focusing mostly on how it relates to you. There may be some other angles worth exploring."
or
"I think there might be a bit more nuance to consider here, and it could help to look at it from a wider perspective beyond personal experience."
> Obnoxious people have repeatedly shown to be detrimental to productivity at the organizational level.
You confused directness and openness with obnoxiousness here. The issue with many orgs is they foster fakeness and beating around the bush in an attempt not to offend the easily offended people. This trend also infected the companies from countries with way more direct culture in an attempt to accommodate people from indirect cultures.
7 replies →
[dead]