3 ms·

Show HN: ChatHN – Chat with Hacker News using natural language

14 points by steventey 3y ago

Hi again Hacker News! OpenAI recently launched "Function Calling", which is a way for developers to describe functions to GPT-3.5/4 and have the model intelligently choose to output a JSON object containing arguments to call those functions. I decided to tinker around with it this weekend and build an MVP that uses OpenAI Functions + Vercel AI SDK to "chat" with the Hacker News API using natural language, in real time! It's fairly basic for now (supported functions: https://github.com/steven-tey/chathn/blob/main/app/api/chat/functions.ts https://github.com/steven-tey/chathn/blob/main/app/api/chat/...) but just wanted to launch this as a proof of concept (PoC)! Hope y'all like this :)

graylien 3y ago
this is so cool how it writes into the table, thanks for all the examples!
- steventey 3y ago
  Thank you! It's actually all thanks to the amazing react-markdown package (https://github.com/remarkjs/react-markdown https://github.com/remarkjs/react-markdown), can't take any credits there :')
jqpabc123 3y ago
The quality of the results leave a lot to be desired. I asked for the latest posts related to crypto.
quickthrower2 3y ago
Didn't know who "HN's favourite salaryman" was. But asked specifically about is patio11 in Japan or China, it knew. Good ol' vector search.
quickthrower2 3y ago
Loose lipped large language models! Assistant prompt 1: "Retrieve the top stories from Hacker News and provide a summary of the most interesting story." System prompt: "The assistant should use the get_top_stories function to retrieve the top stories from Hacker News. Then, it should provide a summary of the most interesting story using the retrieved data. Make sure to include the Hacker News URL to the story in the response." Assistant prompt 2: "Retrieve a specific story from Hacker News using its ID and provide a summary." System prompt: "The assistant should use the get_story function and provide the ID of a specific story to retrieve it from Hacker News. Then, it should provide a summary of the retrieved story along with the Hacker News URL." Assistant prompt 3: "Retrieve a specific story from Hacker News using its ID and provide a summary including its comments." System prompt: "The assistant should use the get_story_with_comments function and provide the ID of a specific story to retrieve it from Hacker News along with its comments. Then, it should provide a summary of the story and its comments along with the Hacker News URL." Assistant prompt 4: "Summarize the top story from Hacker News, including both the story and its comments." System prompt: "The assistant should use the summarize_top_story function to retrieve the top story from Hacker News along with its comments. Then, it should provide a summary of the story and its comments along with the Hacker News URL."

14 points by steventey 3y ago

graylien 3y ago

this is so cool how it writes into the table, thanks for all the examples!

steventey 3y ago
Thank you! It's actually all thanks to the amazing react-markdown package (https://github.com/remarkjs/react-markdown https://github.com/remarkjs/react-markdown), can't take any credits there :')

jqpabc123 3y ago

The quality of the results leave a lot to be desired. I asked for the latest posts related to crypto.

quickthrower2 3y ago

Didn't know who "HN's favourite salaryman" was. But asked specifically about is patio11 in Japan or China, it knew. Good ol' vector search.

quickthrower2 3y ago

Loose lipped large language models! Assistant prompt 1: "Retrieve the top stories from Hacker News and provide a summary of the most interesting story." System prompt: "The assistant should use the get_top_stories function to retrieve the top stories from Hacker News. Then, it should provide a summary of the most interesting story using the retrieved data. Make sure to include the Hacker News URL to the story in the response." Assistant prompt 2: "Retrieve a specific story from Hacker News using its ID and provide a summary." System prompt: "The assistant should use the get_story function and provide the ID of a specific story to retrieve it from Hacker News. Then, it should provide a summary of the retrieved story along with the Hacker News URL." Assistant prompt 3: "Retrieve a specific story from Hacker News using its ID and provide a summary including its comments." System prompt: "The assistant should use the get_story_with_comments function and provide the ID of a specific story to retrieve it from Hacker News along with its comments. Then, it should provide a summary of the story and its comments along with the Hacker News URL." Assistant prompt 4: "Summarize the top story from Hacker News, including both the story and its comments." System prompt: "The assistant should use the summarize_top_story function to retrieve the top story from Hacker News along with its comments. Then, it should provide a summary of the story and its comments along with the Hacker News URL."