←back to thread

Getting AI to write good SQL

(cloud.google.com)
478 points richards | 4 comments | 16 May 25 21:10 UTC | HN request time: 0s | source
Show context
mousetree ◴[16 May 25 22:02 UTC] No.44010169[source]▶
Out of all the AI tools and models I’ve tried, the most disappointing is the Gemini built into BigQuery. Despite having well named columns with good descriptions it consistently gets nowhere close to solving the problem.
replies(2): >>44010326 #>>44010346 #
1. quantadev ◴[16 May 25 22:27 UTC] No.44010326[source]▶
Having proper constraints and foreign keys that are clear is generally all that's needed in my experience. Are you sure your tables have well defined constraints, so that the AI can be absolutely 100% sure how everything links up? SQL is very precise, but only if you're utilizing constraints and foreign key definitions well.
replies(1): >>44010802 #
2. carderne ◴[16 May 25 23:45 UTC] No.44010802[source]▶
It’s BigQuery, so it likely won’t have any of these.
replies(1): >>44011188 #
3. quantadev ◴[17 May 25 00:55 UTC] No.44011188[source]▶
BigQuery supports all those SQL things I mentioned.
replies(1): >>44011214 #
4. carderne ◴[17 May 25 01:02 UTC] No.44011214{3}[source]▶
I’m just saying it’s likely they aren’t using them. But clearly you should if you want LLMs to do anything useful.