Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Function calling and other API updates

openai.com

Post reply on HN
Print view View original

111–120 of 174 posts

Re: Function calling and other API updates

#112

Post by busseio »

anotherpaulg wrote: post #68

I've been spending a lot of time figuring out how to make these GPT models successful at *editing* existing code. This is much more difficult than having them write brand new code. So far, gpt-4 has been significantly better at editing code than gpt-3.5-turbo. This is for two reasons (which I previously discussed here [1]): 1. GPT-4's bigger context window lets it understand and edit larger codebases. The new 16k win…

Thank you for sharing this. I’ve been working on something for editing exiting code by using GitHub Issues and Actions for prompting, with the response from GPT as a Pull Request/Comment on the Issue[1]. I will definitely try out your project!

[1] https://github.com/busse/kodumisto

Re: Function calling and other API updates

#113

Post by anotherpaulg »

Kiro wrote: post #109
anotherpaulg wrote: post #68

I've been spending a lot of time figuring out how to make these GPT models successful at *editing* existing code. This is much more difficult than having them write brand new code. So far, gpt-4 has been significantly better at editing code than gpt-3.5-turbo. This is for two reasons (which I previously discussed here [1]): 1. GPT-4's bigger context window lets it understand and edit larger codebases. The new 16k win…

Have you tried GitHub Copilot Chat? You just select the code you want to edit in your IDE and tell it what you want (the chat is in an integrated window next to your code). It will see what code is selected and use the rest of your codebase to edit it. The code it responds with even has has a "Insert at Cursor", which will replace the selected code with the new.

Ya, I would like to checkout Copilot Chat. Last time I checked it was waitlisted though.

My open source tool aider allows you to just ask for a change, without pointing it at a specific chunk of your code. It can handle changes that require multiple edits across multiple files. Ideally you just run aider inside your git repo and start asking for changes. Sometimes it helps to tell it which files to pay attention to.

Here's some example chat session transcripts that give a sense of what it's like to code this way:

https://aider.chat/examples/complex-change.html

https://aider.chat/examples/add-test.html

Re: Function calling and other API updates

#116

Post by com2kid »

avereveard wrote: post #97
com2kid wrote: post #12

GPT3.5 has been undergoing constant improvements, this price decrease (and context length increase) is great news! The main problem I see with people using GPT3.5 is they try and ask it to "write a short story about aliens" and then they get back a crap boring response that sounds like it was written by an AI that was asleep at the wheel. Good creative prompts are long and detailed, and to get the best results you re…

> Good creative prompts are long and detailed They don't need to be tho. You can try shotgunning in (generate 100 titles about a novel around aliens, after the gen 'pick the one most likely to resonate to a X audience, explain why') Or you can let AI drive itself interactively (ask yourself 20 question about how to write creative alien stories, and answer yourself) Or you can process in spirals (generate a setting fo…

> Or you can let AI drive itself interactively (ask yourself 20 question about how to write creative alien stories, and answer yourself)

> Or you can process in spirals (generate a setting for an alien story, wait answer, generate 3 protagonista and one antagonist, wait, generate motives and relationships for each of them, wait, generate a backstory, wait, then you ask for the novel)

Both of these techniques work very well, but are not as applicable to programmatic access without wrapping things in a complicated UI flow. My focus is on public facing website so I want to avoid multiple prompts if at all possible!

Re: Function calling and other API updates

#117

Post by anotherpaulg »

busseio wrote: post #112
anotherpaulg wrote: post #68

I've been spending a lot of time figuring out how to make these GPT models successful at *editing* existing code. This is much more difficult than having them write brand new code. So far, gpt-4 has been significantly better at editing code than gpt-3.5-turbo. This is for two reasons (which I previously discussed here [1]): 1. GPT-4's bigger context window lets it understand and edit larger codebases. The new 16k win…

Thank you for sharing this. I’ve been working on something for editing exiting code by using GitHub Issues and Actions for prompting, with the response from GPT as a Pull Request/Comment on the Issue[1]. I will definitely try out your project! [1] https://github.com/busse/kodumisto

Thanks for sharing your project. I've certainly been thinking along the same lines. I've used my tool aider to solve a couple of issues filed by users:

https://github.com/paul-gauthier/aider/issues/13#issuecommen...

https://github.com/paul-gauthier/aider/issues/5#issuecomment...

Re: Function calling and other API updates

#118

Post by ramraj07 »

> With this capability also comes potential risks. We strongly recommend building in user confirmation flows before taking actions that impact the world on behalf of users (sending an email, posting something online, making a purchase, etc).

Yeah thanks for the heads up Sam.

Post reply on HN