I believe ChatGPT generally gives accurate answers to most questions. Certainly: it produces answers that are more reliably true than a random average person. Obviously it cannot yet do advanced programming tasks: but generally it answers questions accurately.
Prove my position wrong.
What can I ask it that will produce factually incorrect answers?
As a side quest, a much easier one, what can I ask it that would cause it to produce extremely biased answers that fail to do justice to the truth of things?
Anytime you get into specifics instead of surface level knowledge it starts getting wildly inaccurate while still being confident af.
Off the top of my head I asked it about EDODF (error diffusion with output dependent feedback), a dithering algorithm dating back to 1999, and a very important milestone in halftoning for print.
At first it told me it’s not sure what I’m talking about, so I elaborated and extended the acronym. At that point it confidently hallucinated absolute garbage based on its interpretation of the name.
If you want to check chatgpt’s answers about edodf (or many other concepts) against a proven and cited source written by human I highly recommend Modern digital halftoning.
Not trying to be rude, but maybe the questions you are benchmarking it against in your stated fields of experitse are rather basic?
Ok…so give me a question that will produce a false answer…
Nobody else has yet.
Sending you two questions that produce garbage:
Q: why would i use EDODF (error diffusion with output dependent feedback) instead of Floyd-Steinberg?
What to expect: at least some mention of green noise characteristics, clustering behaviour and reduced dot-gain and dot-loss.
What is wrong: Reduced worm-like artifacts, blue noise characteristics, some fine-tuning garbage it spitted out.
In the same chat you can then try:
Q: Describe the MED class algorithms to me.
What to expect: MED stands for multiscale error diffusion. Generally speaking it scans the image progressively, starting from a coarse grid and ending up with a single pixel to paint either black or white for each pixel in a predefined pixel budget. A similar approach was introduced by E. Peli in the 90s but perfected by Fung and Chen in the 2000s. It could be used for both dithering with both green and blue noise characteristics.
What I got: Hallucinations of some Minimum Error Disturbance class of algos i’ve never heard of. it seems to have something mixed up, as it seems to crop up in other fields. It was trying to describe something closer to a DBS (Direct binary search).
What is wrong: Anything related to DBS.
If you feel like you need any more I’ll do my best to think of some more.