Hacker Newsnew | past | comments | ask | show | jobs | submit | roxyrox's commentslogin

agreed. it’s practically a prerequisite for everything else in the book. Staying calm and thinking clearly is foundational


wow so many warnings for the future.. They didnt intend to but FB now has some responsibility about whats generated on it as one of the most massive source of info in the planet...


Have you ever tried Source-graph ? To handle such consistency issues. (we are trying to do the same at Anyshift for Terraform code) For me the issue is only be exacerbated by gen AI and the era of "big code" thats ahead


Agreed. AI is a great enabler but is becoming a commodity. Now it's more a feature that you leverage based on other integration to have real value on it. I se it as a great feature outside of creatively related tasks.


argh so you had to retest the entire prompting / results ? did you do it exhaustively?


You don't have to redo the whole thing, but there are nuances, just like switching between any framework. I was actually able to use a much shorter prompt with VertexAI than OpenAI.

One of the big problems we will have with these large, hosted models is version changes. They don't support them for very long because they are expensive to run and they only have so much capacity. You typically have to redo prompts with each new version. This is why I think self-hosted will eventually win. The upstream vendor will break your pipelines and there's not much you can do about it. With models you control, you decide when they update and can test / update thoroughly before you do.


first python, but do you have one in mind that's particularly painful?


tks :)


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: