Earlier today Cursor, the magical AI-powered IDE started kicking users off when they logged in from multiple machines.
Like,you’d be working on your desktop, switch to your laptop, and all of a sudden you're forcibly logged out. No warning, no notification, just gone.
Naturally, people thought this was a new policy.
So they asked support.
And here’s where it gets batshit: Cursor has a support email, so users emailed them to find out. The support peson told everyone this was “expected behavior” under their new login policy.
One problem. There was no support team, it was an AI designed to 'mimic human responses'
That answer, totally made up by the bot, spread like wildfire.
Users assumed it was real (because why wouldn’t they? It's their own support system lol), and within hours the community was in revolt. Dozens of users publicly canceled their subscriptions, myself included. Multi-device workflows are table stakes for devs, and if you're going to pull something that disruptive, you'd at least expect a changelog entry or smth.
Nope.
And just as people started comparing notes and figuring out that the story didn’t quite add up… the main Reddit thread got locked. Then deleted. Like, no public resolution, no real response, just silence.
To be clear: this wasn’t an actual policy change, just a backend session bug, and a hallucinated excuse from a support bot that somehow did more damage than the bug itself.
But at that point, it didn’t matter. People were already gone.
Honestly one of the most surreal product screwups I’ve seen in a while. Not because they made a mistake, but because the AI support system invented a lie, and nobody caught it until the userbase imploded.
Sadly, Cursor will always be hampered by maintaining it's own VSCode fork. Others in this niche are expanding rapidly and I, myself, have started transitioning to using Roo and Cline.
https://www.cursor.com/how-to-install-extension
Which is basically an article to use an extension in a way that’s basically forbidden use.
If that was not bad enough the editor also told you to install certain extensions if certain file extensions were used that were also against the tos of the extension.
And basically cursor can just be using the vsix marketplace from eclipse, which does not contain restricted extensions.
What they do is at least shady.
And yes I’m not a fan of the fact that Microsoft does this, even worse they closed the source (or some parts of it) of some extensions as well, which is also a bad move (but their right)
VibeSupport bots, how well did that work out for Canada Air?
https://thehill.com/business/4476307-air-canada-must-pay-ref...
I don't know what the hypothetical penalty would be for mere use contrary to EULA, though. It would be breach of contract, and presumably the court would determine actual damages, but I don't know what cost basis there would be if the software in question was distributed freely. However, fine or no fine, I would expect the court to order the defendant to cease using software in violation of EULA, and at that point further use would be contempt of court, no?
The section about hallucinations is deeply relevant.
Namely, Claude sometimes provides a plausible but incorrect chain-of-thought reasoning when its “true” computational path isn’t available. The model genuinely believes it’s giving a correct reasoning chain, but the interpretability microscope reveals it is constructing symbolic arguments backward from a conclusion.
https://en.wikipedia.org/wiki/On_Bullshit
This empirically confirms the “theory of bullshit” as a category distinct from lying. It suggests that “truth” emerges secondarily to symbolic coherence and plausibility.
This means knowledge itself is fundamentally symbolic-social, not merely correspondence to external fact.
Knowledge emerges from symbolic coherence, linguistic agreement, and social plausibility rather than purely from logical coherence or factual correctness.
Additionally, in the example you share, where only one person knows the context of the change, code review is an excellent tool for knowledge sharing.
[0]: https://dl.acm.org/doi/10.1145/2597073.2597076, for example
Just because you use AI for customer service doesn't mean you don't have to honour its offers to customers. Air Canada recently lost a case where its AI offered a discount to a customer but then refused to offer it "IRL"
https://www.forbes.com/sites/marisagarcia/2024/02/19/what-ai...
So I've always avoided using the Windows Store on my Windows machines, I think I managed to get WSL2 installed without using it lol.
So I'm not sure on the details, but do the steps on https://www.cursor.com/how-to-install-extension bypass clicking "I agree" since they just download and drag? Because from what I can tell, the example in https://www.elgaronline.com/edcollchap/edcoll/9781783479917/... is because the customer clicked "I agree" before installing.
Idk. It’s just growing pains. Companies that grow quickly have problems. Imma keep using https://cline.bot and Claude 3.7.
https://link.springer.com/article/10.1007/s10676-024-09775-5
> # ChatGPT is bullshit
> Recently, there has been considerable interest in large language models: machine learning systems which produce human-like text and dialogue. Applications of these systems have been plagued by persistent inaccuracies in their output; these are often called “AI hallucinations”. We argue that these falsehoods, and the overall activity of large language models, is better understood as bullshit in the sense explored by Frankfurt (On Bullshit, Princeton, 2005): the models are in an important way indifferent to the truth of their outputs. We distinguish two ways in which the models can be said to be bullshitters, and argue that they clearly meet at least one of these definitions. We further argue that describing AI misrepresentations as bullshit is both a more useful and more accurate way of predicting and discussing the behaviour of these systems.
We're embedding "Active Curation" into the workflow: a semi-automated, human-guided loop that refines tickets, PRs, datasets, models, and scripted behaviors in response to real-world feedback. It's a synergistic, self-reinforcing system — every issue flagged by a user can improve detection, drive model updates, shape downstream actions, and tighten the entire product feedback loop across tools and teams.
So consider this tough love, from someone who cares:
Cursor totally missed the boat on the customer support hallucination fiasco. Not just by screwing up the response — that happens — but by failing to turn the whole mess into a golden opportunity to show they understand the limits of LLMs, and how to work with those limits instead of pretending they don’t exist.
They could have said: Here’s how we’re working to build an AI-powered support interface that actually works — not by faking human empathy, but by exposing a well-documented, typed, structured interface to the customer support system.
You know, like Majordomo did 30 years ago, like GitHub did 17 years ago, or like MCP does now — with explicit JSON schemas, embedded documentation, natural language prompts, and a high-bandwidth contract between the LLM and the real world. Set clear expectations. Minimize round trips. Reduce misunderstandings.
Instead? I got ghosted. No ticket number. No public way to track my issue. I wrote to enterprise support asking specifically for a ticket number — so I could route future messages properly and avoid clogging up the wrong inboxes — and got scolded by a bot for not including the very ticket number I was asking for, as if annoyed I'd gone around its back, and being dense and stubborn on purpose.
You play with the Promethean fire of AI impersonating people, that's what you get, is people reading more into it than it really means! It's what Will Wright calls the "Simulator Effect" and "Reverse Over-Engineering".
https://donhopkins.medium.com/designing-user-interfaces-to-s...
Eventually, after being detected trying to get through on the corporate email address, I was pawned off to the hoi polloi hi@cursor.com people-bot instead of the hoi aristoi enterprise@cursor.com business-bot. If that was a bot, it failed. If it was a human, they wrote like a bot. Either way, it's not working.
And yes — the biggest tell it wasn't a bot? It actually took hours to days to respond, especially on weekends and across business hours in different time zones. I literally anthropomorphized the bot ghosting into an understandably overworked work-life-balanced human taking a well earned weekend break, having a sunny poolside barbecue with friends, like in a Perky Pat Layout, too busy with living their best life to answer my simple question: "What is the issue ID you assigned to my case, so we can track your progress?" so I can self serve and provide additional information, without bothering everyone over email. The egg is on my face for being fooled by a customer support bot!
Cursor already integrates deeply with GitHub. Great. They never linked me to any ticketing system, so I assume they don't expose it to the public. That sucks. They should build customer support on top of GitHub issues, with an open-source MCP-style interface. Have an AI assistant that drafts responses, triages issues, suggests fixes, submits PRs (with tests!) — but never touches production or contacts customers without human review. Assist, don’t impersonate. Don’t fake understanding. Don’t pretend LLMs are people.
That’s not just safer — it’s a killer dev experience. Cursor users already vibe-code with wild abandon. Give them modular, extensible support tooling they can vibe-code into their own systems. Give them working plugins. Tickets-as-code. Support flows as JSON schemas. Prompt-driven behaviors with versioned specs. Be the IDE company that shows other companies how to build world-class in-product customer support using your own platform fully integrated with GitHub.
We’re doing this at Leela. We'd love to build on shared open foundations. But Cursor needs to show up — on GitHub, in issue threads, with examples, with tasty dogfood, and with real engineering commitment to community support.
Get your shit together, Cursor. You're sitting on the opportunity of a generation — and we’re rooting for you.
----
The Receipts:
----
Don to Sam, also personally addressed to the enterprise and security bots (explicitly asking for an issue ID, and if it's human or not):
>Hello, Sam.
>You have not followed up on your promise to reply to my issue.
>When will you reply?
>What is the issue ID you assigned to my case, so we can track your progress?
>Are you human or not?
>-Don
----
Enterprise and Security bots: (silence)
Sam to Don (ignoring my request for an issue ID, and my direct question asking it to disclose if it's human or not):
>Hi Don - I can see you have another open conversation about your subscription issues. To ensure we can help you most effectively, please continue the conversation in your original ticket where my teammate is already looking into your case. Opening new tickets won't speed up the process. Thanks for your patience!
----
Don to Sam (thinking: "LLMs are great at analyzing logs, so maybe if I make it look like a cascade of error messages, it will break out of the box and somebody will notice):
>ERROR: I asked you for my ticket number.
>ERROR: I was never given a ticket number.
>ERROR: You should have inferred I did not have a ticket number because I asked you for my ticket number.
>ERROR: You should not have told me to use my ticket number, because you should have known I did not have one.
>ERROR: Your behavior is rude.
>ERROR: Your behavior is callous.
>ERROR: Your behavior is unhelpful.
>ERROR: Your behavior is patronizing.
>ERROR: Your behavior is un-empathic.
>ERROR: Your behavior is unwittingly ironic.
>ERROR: Your behavior is making AI look terrible.
>ERROR: Your behavior is a liability for your company Cursor.
>ERROR: Your behavior is embarrassing to your company Cursor.
>ERROR: Your behavior is losing money for your company Cursor.
>ERROR: Your behavior is causing your company Cursor to lose customers.
>ERROR: Your behavior is undermining the mission of your company Cursor.
>ERROR: Your behavior is detrimental to the success of your company Cursor.
>I would like to speak to a human, please.
----
Four hours and 34 minutes from sending that I finally got a response from a human (or a pretty good simulation), who actually read my email, and started the process of solving my extremely simple and stupid problem, which my initial messages -- if anyone read them or ran a vision model on all the screen snapshots I provided -- would have given them enough information to easily solve the problem in one shot.
It's true that we use tools with uncertainty all the time, in many domains. But crucially that uncertainty is carefully modeled and accounted for.
For example, robots use sensors to make sense of the world around them. These sensors are not 100% accurate, and therefore if the robots rely on these sensors to be correct, they will fail.
So roboticists characterize and calibrate sensors. They attempt to understand how and why they fail, and under what conditions. Then they attempt to cover blind spots by using orthogonal sensing methods. Then they fuse these desperate data into a single belief of the robot's state, which include an estimate of its posterior uncertainty. Accounting for this uncertainty in this way is what keeps planes in the sky, boats afloat, and driverless cars on course.
With LLMs It seems like we are happy to just throw out all this uncertainty modeling and to leave it up to chance. To draw an analogy to robotics, what we should be doing is taking the output from many LLMs, characterizing how wrong they are, and fusing them into a final result, which is provided to the user with a level of confidence attached. Now that is something I can use in an engineering pipeline. That is something that can be used as a foundation to something bigger.
https://old.reddit.com/r/cursor/comments/1jyy5am/psa_cursor_...
(For reference, here it is in reveddit https://www.reveddit.com/v/cursor/comments/1jyy5am/psa_curso... - text from post was unfortunately not saved)
It's already locked and with a stickied comment from a dev clarifying what happened
Did you remove it so people can't find about this screwup when searching Google?
Anyway, if you acknowledge it was a mistake to remove the thread, could you please un-remove it?
But in English it would be just "Capital", right? (The uncountable nouns are rarely used with articles, it's "happiness" not "the happiness". See also https://old.reddit.com/r/writing/comments/12hf5wd/comment/jf... )
https://marginalrevolution.com/marginalrevolution/2017/10/pi...
And then later similar claims about inequality were similarly made using bad methodology (data).
https://marginalrevolution.com/marginalrevolution/2023/12/th...
[1] "Indeed, in some cases, Sutch argues that it has risen more than Piketty claims. Sutch is rather a journeyman of economic history upset not about Piketty’s conclusions but about the methods Piketty used to reach those conclusions."
My company provides hallucination detection software: https://cleanlab.ai/tlm/
But we somehow end up in sales meetings where the person who requested the meeting claims their AI does not hallucinate ...
You can try out our early product here: https://cleanlab.ai/tlm/
(free to try, we'd love to hear your feedback)
