Technology

59377 readers

5811 users here now

This is a most excellent place for technology news and articles.

Our Rules

Follow the lemmy.world rules.
Only tech related content.
Be excellent to each another!
Mod approved content bots can post up to 10 articles per day.
Threads asking for personal tech support may be deleted.
Politics threads may be removed.
No memes allowed as posts, OK to post as comments.
Only approved bots from the list below, to ask if your bot can be added please contact us.
Check for duplicates before posting, duplicates may be removed

Approved Bots

founded 1 year ago

MODERATORS

775

ChatGPT Answers Programming Questions Incorrectly 52% of the Time: Study (gizmodo.com)

submitted 5 months ago by ForgottenFlux@lemmy.world to c/technology@lemmy.world

198 comments fedilink hide all child comments

The research from Purdue University, first spotted by news outlet Futurism, was presented earlier this month at the Computer-Human Interaction Conference in Hawaii and looked at 517 programming questions on Stack Overflow that were then fed to ChatGPT.

“Our analysis shows that 52% of ChatGPT answers contain incorrect information and 77% are verbose,” the new study explained. “Nonetheless, our user study participants still preferred ChatGPT answers 35% of the time due to their comprehensiveness and well-articulated language style.”

Disturbingly, programmers in the study didn’t always catch the mistakes being produced by the AI chatbot.

“However, they also overlooked the misinformation in the ChatGPT answers 39% of the time,” according to the study. “This implies the need to counter misinformation in ChatGPT answers to programming questions and raise awareness of the risks associated with seemingly correct answers.”

you are viewing a single comment's thread
view the rest of the comments

[–] tranxuanthang@lemm.ee 13 points 5 months ago* (last edited 5 months ago) (8 children)

If you don't know what you are doing, and you give it a vague request hoping it will automatically solve your problem, then you will just have to spend even more time to debug its given code.

However, if you know exactly what needs do do, and give it a good prompt, then it will reward you with a very well written code, clean implementation and comments. Consider it an intern or junior developer.

Example of bad prompt: My code won't work [paste the code], I keep having this error [paste the error log], please help me

Example of (reasonably) good prompt: This code introduces deep recursion and can sometimes cause a "maximum stack size exceeded" error in certain cases. Please help me convert it to use a while loop instead.

[–] CapeWearingAeroplane@sopuli.xyz 1 points 5 months ago

I've found chatgpt reasonably good for one thing: Generating regex-patterns. I don't know regex for shit, but if I ask for a pattern described with words, I get a working pattern 9/10 times. It's also a very easy use-case to double check.

load more comments (7 replies)