News for hackers. Stuff that matters.

A comment on Replacing a SQL analyst with 26 recursive GPT prompts. Parent comment

Re:Replacing a SQL analyst with 26 recursive GPT prompts by jeltz on Wednesday January 25, 2023 @12:57PM (#34521720)
Yeah, it doesn't. A friend of mine played around with this and on trickier questions you got confidently incorrect answers.
[ Parent | on HN ]
Re:Replacing a SQL analyst with 26 recursive GPT prompts by LastTrain on Wednesday January 25, 2023 @12:59PM (#34521768)
That is almost worse than it completely not working.
[ Parent | on HN ]
Re:Replacing a SQL analyst with 26 recursive GPT prompts by modriano on Wednesday January 25, 2023 @01:20PM (#34522073)
In my experience, it's unambiguously worse to produce wrong answers confidently than to return no answers, although I expect confidence men (also known as "con men") have a different opinion on this.
[ Parent | on HN ]
Re:Replacing a SQL analyst with 26 recursive GPT prompts by williamcotton on Wednesday January 25, 2023 @03:12PM (#34523513)
Hard code a preface of “My best guess is, and I could be wrong:”?

Show both the SQL query and the results and politely request that the user double check the work? Show multiple generated responses from different seeds?

[ Parent | on HN ]
Re:Replacing a SQL analyst with 26 recursive GPT prompts by pklee on Thursday January 26, 2023 @11:19PM (#34542425)
This is fantastic. This is what we do on our platform. + Copy and keep what worked and what did not so you can fine tune :)
[ Parent | on HN ]
Re:Replacing a SQL analyst with 26 recursive GPT prompts by typon on Wednesday January 25, 2023 @01:46PM (#34522422)
Exactly my experience trying to get GPT-3 to produce SQLITE queries. Anything with a CTE, it gets hopelessly wrong. I was super excited to show off this to my team and after one too many false results I just quietly swept this idea under the rug.
[ Parent | on HN ]