GPT-5, why we aren't adding it yet

So, if you haven't played with it; its seriously impressive. The version we run in our workspace takes like 15 minutes for a response but genuinely has amazed me for its technical capabilities - sometimes even spitting out full folders & pages full of production-grade code.

Here's an issue I see for adding (just for right now):

It does not support temperature movement. Any one who has had appointment booking issues in the past know this is literally the difference between a flawless booking flow and 'going off the rails'. We ALWAYS recommend a temperature of 0.00 to be deterministic with data (super important for booking & knowledge base).
New models are unpredictable. We've all experienced it - new model comes out we want to jump on it ASAP only for it to send binary code to a lead out of the blue and act like nothing happened... We generally want to wait 3-4 weeks before even putting it in beta to release, we still have a week or so.

On a personal note, seriously use it - some people will say "its just same ole same ole" it really is not. It is miles and bus-lengths more intelligent than other models; BUT ONLY ON TECHNICAL TASKS LIKE CODING. Its like talking that super super super smart guy, but socially awkward? Know what I mean?

anyways, we will continue to test and add in future.

Happy tuesday <3

17 comments