[LINK] ChatGPT's odds of getting code questions correct are worse than a coin flip

[Jan Whitaker]

You might blow its circuits. I'm sorry, Dave, but I can't do that right now.

You can tell it to compare and contrast, I think.


On 16/08/2023 10:58 am, Bernard Robertson-Dunn wrote:
> On 15/08/2023 9:04 am, Kim Holburn wrote:
>>
>> >Among other findings, the authors found ChatGPT is more likely to 
>> make conceptual errors than factual ones. "Many answers are incorrect 
>> due to ChatGPT’s incapability to understand the underlying context of 
>> the question being asked," the paper found. 
>
> Suppose you wanted to tell an AI system how to distinguish between 
> right and wrong (if you try it will tell you it has no opinion), 
> correct and incorrect, truth and lies? How would you do it?
>
> Would it work with e.g. far right extremists? MAGA supporters? 
> Politicians?
>