I agree, Shane!
We have close to 90 Kix-based apps in production on over 100,000 computers globally. We write our scripts to a very high standard, and check error status on everything, and STILL get caught with crashes when the environment changes, MS changes the rules (call methods) or an endpoint has some AV that blocks a system call but not the application. Our applications actually write telemetry data back to us if they encounter bad data, and we get notified if an application crashes so we can evaluate and improve the data and error handling.
You can't anticipate every possible condition, so checking the return code of your functions is essential to reliability. If ChatGPT is suggesting otherwise, it isn't as "intelligent" as it might seem. 
_________________________
Actually I
am a Rocket Scientist!
