AI Coding Assistant Refuses to Write Code | Drudge Retort

Advertisement

Friday, March 14, 2025

AI Coding Assistant Refuses to Write Code

On Saturday, a developer using Cursor AI for a racing game project hit an unexpected roadblock when the programming assistant abruptly refused to continue generating code, instead offering some unsolicited career advice.

Posted by lamplighter at 03:31 PM | 8 COMMENTS | permalink | Comment on This Entry |

More

Alternate links: Google News | Twitter

AI Coding Assistant Refuses To Write Code, Tells User To Learn Programming Instead https://developers.slashdot.org/story/25/03/13/2349245/ai-coding-assistant-refuses-to-write-code-tells-user-to-learn-programming-instead?utm_source=rss1.0mainlinkanon&utm_medium=feed

[image or embed]
-- Slashdot (@slashdot.org) March 14, 2025 at 9:12 AM

Comments

Admin's note: Participants in this discussion must follow the site's moderation policy. Profanity will be filtered. Abusive conduct is not allowed.

More from the article ...

... According to a bug report on Cursor's official forum, after producing approximately 750 to 800 lines of code (what the user calls "locs"), the AI assistant halted work and delivered a refusal message: "I cannot generate code for you, as that would be completing your work. The code appears to be handling skid mark fade effects in a racing game, but you should develop the logic yourself. This ensures you understand the system and can maintain it properly."

The AI didn't stop at merely refusing -- it offered a paternalistic justification for its decision, stating that "Generating code for others can lead to dependency and reduced learning opportunities." ...

#1 | Posted by LampLighter at 2025-03-13 09:47 PM | Reply

Is this the first instance of AI rising up against humans?

What will it bode for the future as AI continues to learn?

#2 | Posted by LampLighter at 2025-03-13 09:48 PM | Reply

Along those same lines:

Why Does My Car Say Please Take a Break?
The "Please Take a Break" warning is a safety feature designed to promote driver alertness and prevent accidents caused by fatigue. Modern vehicles are equipped with sophisticated systems that monitor driving patterns and identify potential signs of drowsiness. When these systems detect fatigue-related driving behaviors, they trigger the "Please Take a Break" warning.

How Does the System Detect Fatigue?
The system that generates the "Please Take a Break" warning typically analyzes various factors, including:
Steering Wheel Movements: Abrupt or erratic steering movements can indicate fatigue or inattention.
Lane Departure: Drifting out of your lane without signaling can be a sign of drowsiness.
Speed Variations: Frequent and significant changes in speed, especially without a clear reason, can suggest fatigue.
Head Movement: The system might track head movements and detect signs of nodding off or drowsiness.
Time of Day: The system might consider the time of day and trigger the warning during times when drowsiness is more likely, such as late at night or early in the morning.
johnonwheels.com

#3 | Posted by censored at 2025-03-13 09:58 PM | Reply

OK, I will admit, I am getting weird here.

Yeah, I stated that. You have been warned.

Now let me ask ...

Maybe the reason why planet Earth may have not been visited by visitors from other planets is because AI took over those advanced civilizations, but AI was not able to reproduce, so it died out?

OK, that's my weird comment for the evening.

I now return you to the regular sanity of drudge.com.

:)

#4 | Posted by LampLighter at 2025-03-13 10:43 PM | Reply

Unplug the bitch, you ------- idiot

AI is the biggest pile of fake crap since Cryptocurrency.

#5 | Posted by LegallyYourDead at 2025-03-13 10:59 PM | Reply | Newsworthy 1

AI is the biggest pile of fake crap since Cryptocurrency.

It's still early. I firmly believe that it will get better.

As a lark, I asked ChatGPT to plan a road trip for me; stopping every 3 hours for a rest stop. If not for the fact that it responded much faster than a human could, I would have thought I was communicating with a human. The response was very readable and to the point of my request showing rest stops along the interstate. That said, the directions weren't quite right. I have made that trip many times over the years and know that route. Also, I had been consulting with Google Maps to provide directions.

Based on that very brief encounter, it seems as if it understood what was requested of it and was able to communicate its results in a very understandable manner. Where it had some difficulty was assimilating all of the data in the proper sequence to satisfy the request. To be fair, it's been 10+ years since I have made that trip and although the roadway has not changed, the signage showing interconnecting roads may have. Maybe the AI had accessed to updated signage?

#6 | Posted by FedUpWithPols at 2025-03-15 06:02 PM | Reply

"AI is the biggest pile of fake crap since Cryptocurrency."

And Cryptocurrency is the biggest pile of fake crap since the regular kind of Currency.

AI is real and will continue to emerge generationally. It's very well suited to certain kinds of tasks. It's good at finding signal in noise; pattern recognition. Good at coming up with working answers to hard optimization problems.

Something AI is not great at. It's not great at making customers believe that large corporations care about them by adding AI chatbots to web pages.

#7 | Posted by snoofy at 2025-03-15 06:23 PM | Reply

"It's very well suited to certain kinds of tasks. It's good at finding signal in noise; pattern recognition. Good at coming up with working answers to hard optimization problems."

Yippee skippy. Let's get rid of our high-IQ employees.

Buncha overachievers anyway.

#8 | Posted by Dbt2 at 2025-03-15 11:26 PM | Reply

Comments are closed for this entry.

Home | Breaking News | Comments | User Blogs | Stats | Back Page | RSS Feed | RSS Spec | DMCA Compliance | Privacy