4 ms·

I like to use the Bing AI for that because ChatGPT can sometimes hallucinate while Bing will use its knowledge + stuff from websites which makes it a more relia

I like to use the Bing AI for that because ChatGPT can sometimes hallucinate while Bing will use its knowledge + stuff from websites which makes it a more reliable source for things which should match reality.

reaperman 3y ago
On the contrary, I find all the LLMs (including ChatGPT browsing beta) which search to find answers repeat the same blogspam as everyone else. They save time by parsing the blogspam for me, but it’s still low quality. Phind gets an honorable mention though - if that’s the style you prefer I’d strongly recommend it. ChatGPT does hallucinate all the time, but for “learning a language from scratch” it’s very very very minimal. I can’t recall a single time it gave me a response to “idiomatic way to do x in $LANG” which included hallucinated information. Just can’t back it into a corner trying to get it to customize something that needs rigorous optimization for your use case. Overall, the large innate knowledge base of ChatGPT-4 seems to contain a great model of the modern features of most programming languages and their packages (popular PyPI packages, Spring, Boost, Qt, etc, etc). Once I let it browse the internet, quality goes down and roughly matches Bing and Bard. Phind I think gets around this by having a custom ranking of sources or something. Not sure what their secret sauce is but it mostly works pretty well.
- Shawnj2 3y ago
  I don’t think I’ve ever gotten language syntax wrong but ChatGPT will happily tell you to use functions that don’t exist or use them incorrectly. Also yes Bing sources aren’t necessarily quality sources but they are more aligned to reality than what ChatGPT will give you
  - iraldir 3y ago
    is that gpt-4 (with a paid subscription to chatGPT?) The difference is quite massive when it comes to hallucinations
    - Shawnj2 3y ago
      Less often as the first hit on GPT4 but it will sometimes happen far enough into a thread.
  - flangola7 3y ago
    I really only get hallucinations like that with 3.5

reaperman 3y ago

On the contrary, I find all the LLMs (including ChatGPT browsing beta) which search to find answers repeat the same blogspam as everyone else. They save time by parsing the blogspam for me, but it’s still low quality. Phind gets an honorable mention though - if that’s the style you prefer I’d strongly recommend it. ChatGPT does hallucinate all the time, but for “learning a language from scratch” it’s very very very minimal. I can’t recall a single time it gave me a response to “idiomatic way to do x in $LANG” which included hallucinated information. Just can’t back it into a corner trying to get it to customize something that needs rigorous optimization for your use case. Overall, the large innate knowledge base of ChatGPT-4 seems to contain a great model of the modern features of most programming languages and their packages (popular PyPI packages, Spring, Boost, Qt, etc, etc). Once I let it browse the internet, quality goes down and roughly matches Bing and Bard. Phind I think gets around this by having a custom ranking of sources or something. Not sure what their secret sauce is but it mostly works pretty well.

Shawnj2 3y ago
I don’t think I’ve ever gotten language syntax wrong but ChatGPT will happily tell you to use functions that don’t exist or use them incorrectly. Also yes Bing sources aren’t necessarily quality sources but they are more aligned to reality than what ChatGPT will give you
- iraldir 3y ago
  is that gpt-4 (with a paid subscription to chatGPT?) The difference is quite massive when it comes to hallucinations
  - Shawnj2 3y ago
    Less often as the first hit on GPT4 but it will sometimes happen far enough into a thread.
- flangola7 3y ago
  I really only get hallucinations like that with 3.5