#ChatGPT recommends "nsh_main" for creating a #NuttX NSH Task ... But we use "nsh_consolemain" instead

Article: https://lupyuen.github.io/articles/chatgpt?5#nsh-function

Here's the final code that runs OK on Apache #NuttX RTOS ... After lots of scrubbing on #ChatGPT's code

Article: https://lupyuen.github.io/articles/chatgpt?6#correct-code

#ChatGPT forgets that Apache #NuttX RTOS is based on POSIX ... Here's the better POSIX way to start a task

Article: https://lupyuen.github.io/articles/chatgpt?6#correct-code

#ChatGPT will generate various answers to our question on Apache #NuttX RTOS ... And it might get stuck 🤔

Article: https://lupyuen.github.io/articles/chatgpt?7#other-attempts

Will I use #ChatGPT again? Probably not. I don't think I can afford ChatGPT once they start charging ... Just like GitHub Copilot 🤔

Will #ChatGPT make Teaching obsolete? Not if we teach real things ... The creative way!

Article: https://lupyuen.github.io/articles/chatgpt?88#other-attempts

#ChatGPT will gladly answer questions about Apache #NuttX RTOS! But the answers aren't always correct ... Let's find out why

https://lupyuen.github.io/articles/chatgpt

@lupyuen
I tried ChatGPT a bit and it is actually quite full of crap when it comes to encyclopedic notions, mixing History, fictions and social media like it is the same source. I also asked it for online sources and it created very plausible ones but that never existed in the first place. What a mess, and quite useless on that matter, well, worst than useless since it lies rather than saying "I don't know". A bit like a politician or a philosopher.

@leo @lupyuen It lies convincingly. Absolutely awful. We should be training our kids right away not to trust anything not directly from at least a person, no matter how authoritative or charming :) "But mommy, the roomba said..."