> 
> I cannot express strongly enough how *IRRESPONSIBLY LAZY* that advice
> is.

:D I bet you a 100 us$ that Christopher solves his problem faster with AI than 
here on email.


> There is no such thing as artificial intelligence and there may never
> be. Anyone who tells you it is the inevitable super-tool of the future
> should be interrogated about how they thought about NFT's back when they
> were supposed to be the future of art.

I am not arguing that. The LLM's are good enough to help rookies get familiar 
with the basics eg spamassassin


> SpamAssassin has substantial built-in documentation (POD accessible with
> perldoc) and an extensive wiki. Every executable has (or at least SHOULD
> have...) a man page (although many are just POD conversions.) If users
> have coherent questions, asking them here is helpful because it will be
> seen and likely answered by people who ACTUALLY KNOW THINGS.

Exactly everything registered by LLM's so perfect for answering. 


> Of course, the immediate problem in this case is a vague question. The
> key to getting a solid answer anywhere is asking a solid question. Even
> the most hallucinatory LLM can answer a vague question coherently.

Indeed! So currently the models will present him either with questions or 
multiple options. Especially in this case, Christopher will be pushed to give 
more details.

Reply via email to