[{"data":1,"prerenderedAt":153},["ShallowReactive",2],{"project-36471387":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":28,"sentiment":29,"stance_mix":30,"scored_comments":34,"aspects":35,"strip":38,"post_text":40,"what":41,"said":70,"maker_in_thread":98,"shape":101,"grid":104,"attempts":116,"same_week":117,"maker_shows":145,"maker_hidden":146,"reader_version":147,"card":148},"36471387","ChatHN","Chat with Hacker News using natural language","title","Show HN: ChatHN – Chat with Hacker News using natural language","https:\u002F\u002Fchathn.vercel.app\u002F","chathn.vercel.app","steventey",1687718952,2023,14,5,2012,3109,65,"shown","ai","llm chat interface","web_app",[24,25,26,27],"openai function calling","hacker news","natural language search","chatgpt","quiet",null,{"positive":31,"neutral":32,"mixed":33,"critical":32},2,1,0,4,{"performance":36,"design":37},{"positive":32,"critical":32},{"positive":32,"critical":33},[39,32,15,15,33],3,"Hi again Hacker News! OpenAI recently launched \"Function Calling\", which is a way for developers to describe functions to GPT-3.5\u002F4 and have the model intelligently choose to output a JSON object containing arguments to call those functions.\n\nI 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!\n\nIt's fairly basic for now (supported functions: https:\u002F\u002Fgithub.com\u002Fsteven-tey\u002Fchathn\u002Fblob\u002Fmain\u002Fapp\u002Fapi\u002Fchat\u002Ffunctions.ts) but just wanted to launch this as a proof of concept ",{"problem":42,"audience":29,"pricing":29,"motivation":48,"time_to_build":29,"stack":53,"openness":68,"stage":69,"team_size":68},{"value":43,"source":44,"evidence":45},"Querying Hacker News through natural language instead of browsing","inferred",[46],{"id":47},"post",{"value":49,"source":50,"evidence":51},"Tinker with OpenAI function calling; proof of concept","maker",[52],{"id":47},[54,58,62],{"value":55,"source":50,"evidence":56},"OpenAI Function Calling",[57],{"id":47},{"value":59,"source":50,"evidence":60},"Vercel AI SDK",[61],{"id":47},{"value":63,"source":50,"evidence":64},"react-markdown",[65],{"id":66,"by":11,"text":67},"36471537","Thank you! It's actually all thanks to the amazing react-markdown package (https:\u002F\u002Fgithub.com\u002Fremarkjs\u002Freact-markdown), can't take any credits there :')","unknown","prototype",{"praise":71,"objections":87,"requests":95,"questions":96,"alternatives":97},[72,80],{"value":73,"source":74,"evidence":75},"Table output display is cool","commenter",[76],{"id":77,"by":78,"text":79},"36471480","graylien","this is so cool how it writes into the table, thanks for all the examples!",{"value":81,"source":74,"evidence":82},"Vector search answers about people were accurate",[83],{"id":84,"by":85,"text":86},"36477821","quickthrower2","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.",[88],{"value":89,"source":74,"evidence":90},"Quality of results is poor (e.g. crypto query)",[91],{"id":92,"by":93,"text":94},"36473153","jqpabc123","The quality of the results leave a lot to be desired. I asked for the latest posts related to crypto.",[],[],[],{"comments":32,"answered":33,"key":99},[100],{"id":66,"text":67},{"commenters":34,"max_depth":31,"maker_replies":32,"first_comment_mins":102,"lifespan_hours":103},11,16,[105,108,110,112,114],[106,33,39,107,85,33],"36478006","project\u002Fneutral\u002Fnone",[92,32,32,109,93,33],"project\u002Fcritical\u002Fperformance",[84,31,15,111,85,33],"project\u002Fpositive\u002Fperformance",[77,39,15,113,78,33],"project\u002Fpositive\u002Fdesign",[66,39,33,115,11,32],"maker\u002Fneutral\u002Fnone",[],[118,123,128,132,136,140],{"id":119,"name":120,"points":121,"category":122},"36480687","Mofi",708,"creative",{"id":124,"name":125,"points":126,"category":127},"36470297","OpenResume",656,"productivity",{"id":129,"name":130,"points":131,"category":127},"36446045","Vrite Editor",460,{"id":133,"name":134,"points":135,"category":20},"36505859","background removal",395,{"id":137,"name":138,"points":139,"category":20},"36472854","WinGPT",382,{"id":141,"name":142,"points":143,"category":144},"36447871","React Graph Gallery",297,"dev_tools",10,false,"v2",{"id":149,"name":5,"line":6,"ai":146,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":29,"top":18,"kw":150,"maker":11,"domain":10,"tier":19,"attempts":33,"strip":151,"url":9,"mix":152,"praise":73,"objection":89,"sub":21},36471387,[24,25,26,27],"31550",[31,32,33,32],1793991068230]