2 Comments

Great post Christian. It seems like for some orgs (like my own), as soon as there's a comfort level in being able to use the prompt-friendly tech on our own data we could see a massive jump for our internal BI/analytics departments. I think it's one part not wanting to bulk up staffing on the roles can do it with more complex tech and a second part not feeling comfortable just dropping years of business data into models where we have no idea who's using them and for what.

Expand full comment

Thanks Mike. I agree that there's a major comfort-level component to BI/analytics adoption. The obstacle to adoption that I see for BI use-cases is that it's so much harder for the average user to validate whether the results the machine spits out are correct. It's *relatively* easy for me to decide whether a qualitative response from GPT makes sense. But because of the idiosyncracies of companies' internal data, unless I'm restricting what the machine is analyzing to a specific table my data team has cleansed and fool-proofed for me, I have no way of knowing whether the arithmetic was performed on the right fields, nor do I have any prior knowledge of how to filter the results (in effect the WHERE clauses in a SQL query). I do think it's solvable, but I haven't seen a solution for this yet. I'd love to be proven wrong on this if you've seen any solutions that get at this problem!

Expand full comment