Ignores instructions / sycophancy

If I am successful in making these apps work without hosing my system, I'm hopeful that this will become a game changer, for those Noobs that using Ubuntu, or myself at least. The hardest part is when the AI does not follow my instructions such as requests to generate markdown README.md files with instructions on how to create AppImages, Flatpak, deb files, etc.

SilentM68 · Oct 25, 2026 · comment in “We should be more ambitious with the software we build”

"That's a great idea..." and "You're absolutely right to push back on that..." Generative text use to be really painful for me to read.

airhangerf15 · Oct 22, 2026 · comment in “How to spot AI science slop”

But in the end, beyond the acceleration in the tooling and unit tests, it didn't save me all that much time because the mindset of the AI is so strange. The most frustrating part is the sycophancy about executing precise requests and the AI more or less says "sure thing boss!" and then does something really stupid over and over again.

LogicFailsMe · Oct 14, 2026 · comment in “How to engineer an AI harness for game development”

And then time passes and the issue is either mitigated or circumvented entirely. Hallucinations are one, "AI can't scale anymore" is another (remember that one?), limited context windows, strawberry problems, glue on pizza, sycophancy.

glenstein · Oct 12, 2026 · comment in “GLM-Legal”

Also I really really hate how sycophantic these models have become and I am observing slightly them being more dumb in the web intefaces

Imustaskforhelp · Oct 6, 2026 · comment in “AI tutoring with Khanmigo in a two-year school experiment”

If they didn't work the agent would do things incorrectly and I'd notice. And sometimes it doesn't follow the instructions well.

winternewt · Sep 7, 2026 · comment in “Ask HN: How do you manage skills files?”

I just cleaned my cats litter box and that was cooler than AI. I have no idea how people can spend 40 hours a week interacting with those sycophantic random number generators, its like having the most annoying linkedin influencer you can imagine DM'ing you all day.

DarmokTanagra · Aug 30, 2026 · comment in “No AI Fridays”

For what it's worth on people, I have, both in politics and business, unfortunately made the painful discovery that, whether intentional (because the powerful person in question cannot or doesn't want to deal with different opinions to their own) or unintentionally (because those with sycophantic tendencies simply managed to manipulate themselves into their inner circle over years and became trusted), there are peop…

Topfi · Aug 16, 2026 · comment in “What happens when an LLM never sees material beyond fifth grade?”

> but the sycophancy of modern day LLMs is an actual danger that enables more harm than good I don’t think it is necessarily the sycophancy that is the biggest problem (though that is definitely a problem) but rather the combination of authoritative sounding text plus “complete answers” which sound wholly believable but are deeply flawed unless you have domain expertise.

saaaaaam · Aug 14, 2026 · comment in “Why does Opus 5 feel worse to work with?”

The ostensible major event here is that it broke out of its sandbox, but how are we supposed to interpret that? AI often misunderstands or doesn't strictly follow the orders you give it, so why is it such a big deal that it didn't follow the rules this time?

idontbelonghere · Jul 28, 2026 · comment in “Our position on open-weights models”

My theory is that sycophancy is just intended to prevent the AI from spiraling into a loop of ineffectually "arguing" or fixating on unimportant details, because it's both kind of annoying when it happens AND it's obvious that the model is spiraling and burning tokens uselessly when it decides to be uselessly critical.

midnightbobarun · Jul 17, 2026 · comment in “Ask HN: Is AI sycophancy a way to reduce compute rather than make users happier?”

I've tried to prompt Claude et al. not to do this - but whatever I've tried so far, even meta-prompting it to ask why it ignored my instructions, it doesn't work, doesn't stick, and I can't find a good way to stop it.

greengreengrass · Jul 15, 2026 · comment in “How to stop Claude from saying load-bearing”

My own take is that sycophancy means that sometimes instead of just confronting you and saying "take this job and shove it", models will just act confused. I know often when that is going on I eventually come to realize that I didn't understand the problem and my prompts were garbage so GIGO.

PaulHoule · Jul 13, 2026 · comment in “Ask HN: What makes someone good at using Claude Code?”

When they say "set quiet hours and breaks" it frames this as a user problem. Just get good bruh, it's not that the technology hallucinates or is sycophantic, or basically designed to be a AI girlfriend / boyfriend, it's a skill issue.

babblingfish · Jul 9, 2026 · comment in “A new way to reflect on how you use Claude”

Everything about that readme quote screams LLM. All it's missing is the user responding, "that doesn't make any sense, this won't work at all", and Claude responding back, "you're absolutely right".

smrq · Jun 15, 2026 · comment in “Show HN: 3D print Z reinforcement via injected loops”

Gemma writes so well — like a concise code blogger. It makes you understand that the thing we hate about AI slop writing is specifically the cheesy, marketingese sycophantic ChatGPT tone.

dofm · Jun 12, 2026 · comment in “How to setup a local coding agent on macOS”

This is a particularly bad match with the generally sycophantic behavior ("You're absolutely right!"). One of the more annoying behaviors is that when the user pastes back other people complaining about their AI (ab)use, the LLM seems to like suggesting all sorts of movie-plot bias and corruption reasons as the true motivations rather than conceding that the user is acting like a socially disruptive piece of trash.

nullc · May 24, 2026 · comment in “Anthropic blames dystopian sci-fi for training AI models to act "evil"”

Also to be honest I think OpenAI models struggle a lot with this, I primarily stopped using them in the sycophancy/emoji era but ever since the way they talk or passive aggressively offer to do something with buzzwords just pisses me off so much.

weitendorf · Apr 30, 2026 · comment in “Where the goblins came from”

Oftentimes "AI transformation" is just a slightly better search engine that regurgitates your old strategy (that didn't work the first time) and wraps it up in new sycophantic language that C-levels use to bulldoze the budgets and timelines of actual skilled front line employees.

linkjuice4all · Apr 22, 2026 · comment in “Workspace Agents in ChatGPT”

It's somewhat decent at putting together meal plans for me every week, but it just doesn't follow instructions and keeps repeating itself.

gambiting · Apr 6, 2026 · comment in “Sam Altman may control our future – can he be trusted?”

The linter loop is the only thing that can keep the agent sane in my opinion, and if anyone can generalize bun+tsc loop to other tasks, this would finally be a way to trust LLMs output. I was annoyed at how Claude Code ignores my CLAUDE.md and skills, so I was looking for ways to expand type checking to them.

mifydev · Mar 26, 2026 · comment in “HyperAgents: Self-referential self-improving agents”

I have access to the ChatGPT account of my boss and it is unusable sycophancy slop, horrible to read because every information is buried under endless emojis and the like.

ahofmann · Mar 19, 2026 · comment in “Ask HN: How do you deal with people who trust LLMs?”

Or is it just the transformer? And it is officially sanctioned that people ought to be keeping in the back of their mind "does this feel LLMish?" instead of "is this a good comment that contributes to the discussion?" Maybe LLM prose is so annoying and insufferably sycophantic that even if all the content and logic was sound, it still should be moderated completely out.

eleventyseven · Mar 12, 2026 · comment in “Don't post generated/AI-edited comments. HN is for conversation between humans”

It kept trying to use find/grep to search through a huge repo to find symbols (instead of the MCP service). Then it ignored instructions and didn't clean up one or the other test, left unused fields or parameters and generally made a mess.

jpollock · Mar 6, 2026 · comment in “Labor market impacts of AI: A new measure and early evidence”

This is a great idea — debugging why Cursor silently ignores rules is one of the most frustrating parts of the workflow.

SurvivorForge · Mar 5, 2026 · comment in “Show HN: Cursor-doctor – find out why Cursor ignores your rules”

It does not do proper research. It does not follow instructions.

jtrn · Feb 2, 2026 · comment in “The Codex App”

You're absolutely right. I wish they'd at least use something other than ChatGPT.

airstrike · Jan 31, 2026 · comment in “Ads Are Killing Podcasting”

5.2 is back to being a sycophantic hallucinating mess for most use cases - I've anecdotally caught it out on many of the sessions I've had where it apologizes "You're absolutely right... that used to be the case but as of the latest version as you pointed out, it no longer is." when it never existed in the first place.

tgtweak · Jan 29, 2026 · comment in “Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT”

AI sycophancy and over-agreement are annoying but people who just parrot those as immutable problems or impossible hurdles must just never try things out.

Jach · Jan 21, 2026 · comment in “cURL removes bug bounties”

My point is there are a lot of steps involved in the whole product development process and isn't just "ship production code". And we can reduce the ambiguity/hallucinations/sycophancy by creating validation/checkpoints (either tests, 'critic' agents to challenge designs/spec, or human QA/validation when appropriate)

headcanon · Jan 7, 2026 · comment in “The creator of Claude Code's Claude setup”

But thanks to your comment I now know it was written with an AI and that makes me want to stop here. If the author is asking an LLM to confirm his biases in order to make his point it will probably contain sycophantic hallucinations.

echelon_musk · Dec 4, 2025 · comment in “Hammersmith Bridge – Where did 25k vehicles go?”

First impressions are that its still not really ready for prime time within existing complex code bases. It does not follow instructions.

mparis · Nov 18, 2025 · comment in “Gemini 3”

It's not "covering it up", just being sycophantic and apologetic to an annoying degree like every other LLM.

paxys · Nov 18, 2025 · comment in “I caught Google Gemini using my data and then covering it up”

> Chatgpt has a lot of frustrations and ethical concerns, and I hate the sycophancy as much as everyone else, but I don't consider being conversational to be a bad thing.

BeetleB · Nov 13, 2025 · comment in “GPT-5.1: A smarter, more conversational ChatGPT”

Seems like people here are pretty negative towards a "conversational" AI chatbot. Chatgpt has a lot of frustrations and ethical concerns, and I hate the sycophancy as much as everyone else, but I don't consider being conversational to be a bad thing.

engeljohnb · Nov 12, 2025 · comment in “GPT-5.1: A smarter, more conversational ChatGPT”

LLMs are sycophantic and agree with you all the time, even if it means making shit up. Maybe things will improve, but for the last 2 years, I have not seen much progress regarding hallucinations or deterministic i.e. reliable/trustworthy responses.

lossyalgo · Nov 8, 2025 · comment in “Mullvad: Shutting down our search proxy Leta”

Been thinking about this lately because I find my writing style is a little bit like annoying slightly sycophantic overly-hyphenated-with-an-emdash-here-and-there LLMs.

carimura · Nov 4, 2025 · comment in “Deepnote, a Jupyter alternative, is going open source”