11 ms·

I've spent a lot of time on this problem space for our business. My current conclusions are as follows: Writing the actual SQL query is not the hard part for u

I've spent a lot of time on this problem space for our business. My current conclusions are as follows: Writing the actual SQL query is not the hard part for us. The hardest part is exhaustively expressing our intention in any language at all, followed closely by its apprentice - the incomplete requirements monster. I've found that if I can coach a team member into giving me a proper natural language statement that we would use as an ideal training example for a SQL bot, then they are probably able to just go ahead and write the final query by hand. Overall, it would likely take less time & cognitive overhead. Don't get me wrong - I see the automation potential too. I've used it for lots of really crappy bulk refactors and it's great. But, it absolutely cannot think for you and magic-in the requirements that your customer ultimately left on the floor. Generative SQL output is the opposite of what you want when you are at the frontier of understanding and incomplete requirements. For us, this is kind of where the biggest use case would have come from. We don't need much help with the boring stuff. We need a really smart asshole to call us out when there appears to be a logical gap somewhere. ChatGPT almost seemed to be this @ launch but I think its been RLHF'd into a potato by now. I stopped using it for anything serious around late May.

samstave 3y ago
>>*...Anything serious...* My problem is that I simply cannot trust any data coming from GPT - it doesnt have real-time info, it constantly tells me what it CANT DO and its never able to tell me what it CAN DO. The platform is extremely powerful, and what we are seeing is that the cadre of AI LLMs/apps that are coming about has all the devs/companies frantically running in a panic on how to both weaponize and monetize.
- nonethewiser 3y ago
  That issue seems entirely mitigated by thorough testing.
- FpUser 3y ago
  >"My problem is that I simply cannot trust any data coming from GPT" I just used it to write shell function which extracts value from .ini file using section, name as a parameters. Works perfectly. Otherwise I would have to manually dick with awk input string wrecking my brain in the process
  - samstave 3y ago
    function validation is not same as information validation and links to information from [world]\\ I would EXPECT/DEMAND your results - other results are not functions.
majormajor 3y ago
Subqueries and windowing and partitioning are annoying tedious areas if you only use that sort of SQL occasionally (<1/month) and the GPT4 API seems to massively help for turning plain text into the appropriate stuff. I know enough to be able to be confident in it after running some tests of the query, but not enough to knock it out cold without a few minutes research.

samstave 3y ago

>>*...Anything serious...* My problem is that I simply cannot trust any data coming from GPT - it doesnt have real-time info, it constantly tells me what it CANT DO and its never able to tell me what it CAN DO. The platform is extremely powerful, and what we are seeing is that the cadre of AI LLMs/apps that are coming about has all the devs/companies frantically running in a panic on how to both weaponize and monetize.

nonethewiser 3y ago
That issue seems entirely mitigated by thorough testing.
FpUser 3y ago
>"My problem is that I simply cannot trust any data coming from GPT" I just used it to write shell function which extracts value from .ini file using section, name as a parameters. Works perfectly. Otherwise I would have to manually dick with awk input string wrecking my brain in the process
- samstave 3y ago
  function validation is not same as information validation and links to information from [world]\\ I would EXPECT/DEMAND your results - other results are not functions.

majormajor 3y ago

Subqueries and windowing and partitioning are annoying tedious areas if you only use that sort of SQL occasionally (<1/month) and the GPT4 API seems to massively help for turning plain text into the appropriate stuff. I know enough to be able to be confident in it after running some tests of the query, but not enough to knock it out cold without a few minutes research.