[{"data":1,"prerenderedAt":616},["ShallowReactive",2],{"project-34897773":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":7,"subcategory":20,"format":21,"keywords":22,"reception":28,"sentiment":29,"stance_mix":30,"scored_comments":35,"aspects":36,"strip":48,"post_text":49,"what":50,"said":84,"maker_in_thread":184,"shape":186,"grid":191,"attempts":577,"same_week":578,"maker_shows":18,"maker_hidden":607,"reader_version":608,"card":609},"34897773","AskHN","Asks GPT-3 questions answered from Hacker News comment threads","ai","Show HN: AskHN","https:\u002F\u002Fwww.patterns.app\u002Fblog\u002F2023\u002F02\u002F19\u002Fask-hn-gpt-embeddings-question-answering\u002F","patterns.app","kvh",1677082171,2023,612,127,36,3109,2,"legend","question answering bot","other",[23,24,25,26,27],"gpt-3","hacker news","semantic search","embeddings","question answering","curious",-0.119,{"positive":31,"neutral":32,"mixed":33,"critical":34},9,38,1,21,69,{"execution":37,"design":39,"docs":41,"idea":43,"privacy":46,"pricing":47},{"positive":18,"critical":38},7,{"positive":18,"critical":40},0,{"positive":40,"critical":42},3,{"positive":44,"critical":45},4,5,{"positive":40,"critical":45},{"positive":40,"critical":40},[33,40,40,40,40,18,33,42,40,40,42,42,18,42,42,44,33,40,40,42,33,44,42,45,18,33,33,42,42,33,42,42,33,44,42,42,42,33,44,33],null,{"problem":51,"audience":49,"pricing":49,"motivation":49,"time_to_build":49,"stack":59,"openness":82,"stage":83,"team_size":82},{"value":52,"source":53,"evidence":54},"Answers questions using collective wisdom from Hacker News threads","commenter",[55],{"id":56,"by":57,"text":58},"34900121","btbuildem","> “AskHN” is a GPT-3 bot I trained on a corpus of over 6.5 million Hacker News comments to represent the collective wisdom of the HN community in a single bot. First sentence of…",[60,68,75],{"value":61,"source":53,"evidence":62},"GPT-3",[63,64],{"id":56,"by":57,"text":58},{"id":65,"by":66,"text":67},"34900306","agolio","Scroll a bit further down and you will see > 7. Put top matching content into a prompt and ask GPT-3 to summarize > 8. Return summary along with direct links to comments back to…",{"value":69,"source":53,"evidence":70},"OpenAI text-embedding-ada-002",[71],{"id":72,"by":73,"text":74},"34898347","neoromantique","I'm sorry to be THAT GUY, but it is addressed in the article :) >GPT embeddings To index these stories, I loaded up to 2000 tokens worth of comment text (ordered by score, max…",{"value":76,"source":53,"evidence":77},"Pinecone",[78],{"id":79,"by":80,"text":81},"34899498","PetrBrzyBrzek","Hi, thanks for the interesting article. I have a question about Pinecone. What is the cost of storing all these vectors?","unknown","launched",{"praise":85,"objections":104,"requests":129,"questions":144,"alternatives":169},[86,97],{"value":87,"source":53,"evidence":88},"Useful for finding related HN posts and includes sources",[89,93],{"id":90,"by":91,"text":92},"34901123","gnicholas","Love that it includes sources — this makes it much more valuable because you can tell if it's giving useful information or just blowing smoke.",{"id":94,"by":95,"text":96},"34938690","wastu","Instead of being summarize tool, this bot is really useful if you want to search related hn post based on abstract thinking imo.",{"value":98,"source":53,"evidence":99},"Pleasantly surprised by answer quality",[100],{"id":101,"by":102,"text":103},"34902440","flappysolutions","I'm quite pleasantly surprised by this response: Question: what are the biggest fallacies of people that post on hackernews? Answer: One of the biggest fallacies of people that…",[105,112,122],{"value":106,"source":53,"evidence":107},"Bot writes in generic GPT style, not HN voice",[108],{"id":109,"by":110,"text":111},"34898358","have_faith","It seems to write in the generic \"style\" of GPT, instead of in the style I would recognise as a HN poster. Is that because of something baked into how the training process works?…",{"value":113,"source":53,"evidence":114},"Description of training is misleading; it summarizes search results rather than being trained on HN",[115,119],{"id":116,"by":117,"text":118},"34901182","stnmtn","I agree, that language could be very improved. This is not a GPT-like LLM whose training corpus is HN comments, which I found to be an extremely interesting idea. Instead, it…",{"id":120,"by":57,"text":121},"34900652","Ah got it. Perhaps they should edit the intro then, it's misleading.",{"value":123,"source":53,"evidence":124},"Answers contain nonsense that sounds plausible",[125],{"id":126,"by":127,"text":128},"34902174","setgree","As is often true of GPT responses, there's some nonsense interspersed here, e.g. the claim that R has \"a more mature package universe\" than Python. I think this is false, but if…",[130,137],{"value":131,"source":53,"evidence":132},"Writeup on specializing an LLM with HN data",[133],{"id":134,"by":135,"text":136},"34898402","georgelyon","Am I correct in understanding that this doesn't actually generate answers based on HN, but instead finds semantically-near comments and sends them verbatim to GPT to summarize?…",{"value":138,"source":53,"evidence":139},"Clarify cost of embedding comments",[140],{"id":141,"by":142,"text":143},"34902839","leobg","I like the project. Had been wanting to do this myself for a long time, because HN has become the first place I go to nowadays for answers, and I value the intelligence and…",[145,152,162],{"value":146,"source":53,"evidence":147},"Is user data opted out of use",[148],{"id":149,"by":150,"text":151},"34898351","egypturnash","Is there a way to opt out of one's comments being used for this?",{"value":153,"source":53,"evidence":154},"Is this fine-tuned or standard GPT-3 API",[155,159],{"id":156,"by":157,"text":158},"34907673","underlines","Just to be sure: This is NOT a finetuned GTP model, but rather standard GPT-3 API, used to summarize search results of a HN Comments DB, based on user input. Right?",{"id":160,"by":57,"text":161},"34900160","My own experiments made me think that the impact of finetuning is comparable to that of a molecule in a drop in a bucket. > “AskHN” is a GPT-3 bot I trained on a corpus of over…",{"value":163,"source":53,"evidence":164},"Who pays the API costs",[165],{"id":166,"by":167,"text":168},"34902810","jaequery","Seeing a ton of projects utilizing chatgpt nowadays. Are the project owners basically paying the API costs out of pocket ? Think it would add up pretty quick especially if front…",[170,177],{"value":171,"source":53,"evidence":172},"Algolia HN search",[173],{"id":174,"by":175,"text":176},"34900696","dalmo3","This is nice! The official algolia search is useless. Otoh, did I miss something or is it only on discord?",{"value":178,"source":53,"evidence":179},"Appending reddit or hackernews to Google searches",[180],{"id":181,"by":182,"text":183},"34898363","MikeTheRocker","I love this! I used to append \"reddit\" to my Google search queries to get best results, but the quality of dialog over there has really dropped in recent years. These days I've…",{"comments":40,"answered":40,"key":185},[],{"commenters":187,"max_depth":188,"maker_replies":40,"first_comment_mins":189,"lifespan_hours":190},110,8,25,101,[192,194,198,200,203,206,210,213,216,219,222,224,227,230,233,236,238,240,244,246,248,249,250,252,256,260,263,266,269,272,275,278,281,284,287,290,293,294,297,300,303,306,309,311,314,317,320,323,326,327,330,334,336,339,343,346,349,352,355,358,360,363,366,370,372,375,378,380,382,385,388,391,394,397,399,401,403,408,411,413,417,421,424,427,430,433,435,437,440,443,446,450,454,457,461,464,468,472,474,476,479,483,486,489,494,498,501,503,505,507,511,514,517,520,524,528,532,534,537,541,544,548,551,555,559,563,566,570,573],[126,40,33,193,127,40],"project\u002Fcritical\u002Fexecution",[195,40,40,196,197,40],"34902651","commenter\u002Fneutral\u002Fnone","leereeves",[199,40,40,196,127,40],"34902749",[201,40,40,196,202,40],"34903883","ux-app",[204,40,40,196,205,40],"34910496","LtdJorge",[207,40,40,208,209,40],"34907681","other\u002Fneutral\u002Fnone","2020aj",[211,40,40,208,212,40],"34903825","shanusmagnus",[214,40,40,208,215,40],"34909304","nine_k",[217,40,40,208,218,40],"34904710","soheil",[220,40,40,196,221,40],"34905664","wisemang",[223,40,40,208,127,40],"34905746",[225,40,40,208,226,40],"34907053","gnulinux",[228,40,40,196,229,40],"34907465","seanhunter",[231,40,40,208,232,40],"34905021","meken",[234,40,40,208,235,40],"34911684","q1w2",[237,40,40,208,127,40],"34915117",[109,33,42,239,110,40],"project\u002Fneutral\u002Fdesign",[241,33,42,242,243,40],"34898565","project\u002Fneutral\u002Fnone","sebzim4500",[56,33,33,245,57,40],"project\u002Fcritical\u002Fdocs",[65,33,42,247,66,40],"project\u002Fneutral\u002Fdocs",[120,33,33,245,57,40],[116,33,33,245,117,40],[251,33,40,208,66,40],"34902716",[253,33,40,254,255,40],"34919041","other\u002Fneutral\u002Fidea","OkGoDoIt",[257,33,42,258,259,40],"34949341","project\u002Fneutral\u002Fidea","efreak",[261,33,40,208,262,40],"34900801","rpastuszak",[264,33,40,208,265,40],"34899166","britzkopf",[267,33,40,208,268,40],"34900425","bradwood",[270,33,40,208,271,40],"34901102","vidarh",[273,33,40,208,274,40],"34901598","actually_a_dog",[276,33,40,208,277,40],"34901961","genericone",[279,33,40,208,280,40],"34902064","Jimmc414",[282,33,40,208,283,40],"34903614","ysavir",[285,33,42,258,286,40],"34900539","clark-kent",[288,33,40,208,289,40],"34900977","reacharavindh",[291,33,40,208,292,40],"34901449","cookie_monsta",[134,18,42,258,135,40],[295,18,42,258,296,40],"34898530","serjester",[298,18,42,242,299,40],"34901097","jerpint",[301,18,42,258,302,40],"34902053","ilaksh",[304,18,33,193,305,40],"34902417","redox99",[307,18,40,208,308,40],"34904472","moremetadata",[101,42,45,310,102,40],"project\u002Fpositive\u002Fexecution",[312,42,33,193,313,40],"34903037","ducktective",[315,42,40,208,316,40],"34905453","autoexec",[318,44,42,258,319,40],"34898172","LeoPanthera",[321,44,42,258,322,40],"34898233","retube",[324,44,42,258,325,40],"34898311","MuffinFlavored",[72,44,42,247,73,40],[328,44,42,242,329,40],"34898354","nkozyra",[331,44,45,332,333,40],"34898685","project\u002Fpositive\u002Fidea","gorbypark",[335,44,42,258,329,40],"34898344",[337,44,40,208,338,40],"34898419","andai",[340,44,33,341,342,40],"34898369","project\u002Fcritical\u002Fnone","bityard",[344,44,40,208,345,40],"34898632","Aromasin",[347,44,40,208,348,40],"34899048","jb1991",[350,44,40,208,351,40],"34901631","heleninboodler",[353,44,40,208,354,40],"34904133","GreenWatermelon",[356,44,40,196,357,40],"34907752","rocho",[359,45,40,208,299,40],"34901074",[361,45,40,208,362,40],"34903204","freeqaz",[364,45,40,208,365,40],"34904168","jn2clark",[367,368,42,242,369,40],"34901391",6,"boywitharupee",[160,38,42,371,57,40],"project\u002Fneutral\u002Fexecution",[373,38,33,193,374,40],"34900325","danuker",[376,38,33,193,377,40],"34902781","notahacker",[379,38,40,196,374,40],"34903721",[381,38,42,371,286,40],"34900605",[383,38,45,384,57,40],"34901470","project\u002Fpositive\u002Fdesign",[386,188,42,258,387,40],"34899601","SilverBirch",[389,188,42,258,390,40],"34899948","inportb",[392,31,40,208,393,40],"34904505","fleddr",[395,31,40,196,396,40],"34905290","Sai_",[398,31,40,196,393,40],"34909840",[181,400,45,332,182,40],10,[402,400,40,196,142,40],"34902540",[404,405,45,406,407,40],"34898518",11,"maker\u002Fpositive\u002Fnone","monkeydust",[149,409,42,410,150,40],12,"project\u002Fneutral\u002Fprivacy",[412,409,42,410,338,40],"34898368",[414,409,33,415,416,40],"34898372","project\u002Fcritical\u002Fprivacy","hombre_fatal",[418,409,40,419,420,40],"34901475","commenter\u002Fneutral\u002Fprivacy","samstave",[422,409,40,419,423,40],"34898615","jdthedisciple",[425,409,33,415,426,40],"34898897","anaganisk",[428,409,33,415,429,40],"34898982","olivierduval",[431,409,40,432,426,40],"34899080","other\u002Fneutral\u002Fprivacy",[434,409,40,419,316,40],"34905247",[436,409,33,415,429,40],"34898949",[438,409,40,208,439,40],"34899988","Minor49er",[441,409,40,208,442,40],"34901137","triyambakam",[79,444,42,445,80,40],13,"project\u002Fneutral\u002Fpricing",[447,448,40,208,449,40],"34906411",14,"kerogerokero",[451,452,42,258,453,40],"34899162",15,"la64710",[455,452,40,208,456,40],"34900961","bayan1234",[458,459,33,193,460,40],"34901698",16,"renewiltord",[141,462,42,463,142,40],17,"project\u002Fmixed\u002Fpricing",[465,466,40,208,467,40],"34899293",18,"motohagiography",[469,470,33,415,471,40],"34908002",19,"badrabbit",[90,473,45,384,91,40],20,[475,34,42,371,286,40],"34900821",[477,34,42,371,478,40],"34900883","gk1",[480,481,42,258,482,40],"34901412",22,"DanielBMarkham",[484,481,40,254,485,40],"34901647","mittermayr",[487,481,42,371,488,40],"34905481","ed_mercer",[490,491,33,492,493,40],"34901686",23,"project\u002Fcritical\u002Fidea","einpoklum",[495,496,40,208,497,40],"34900981",24,"osigurdson",[499,189,42,258,500,40],"34901183","ada1981",[166,502,42,445,167,40],26,[94,504,45,332,95,40],27,[156,506,42,371,157,40],28,[508,509,42,258,510,40],"34900596",29,"tosemlep",[512,513,42,371,325,40],"34898300",30,[515,513,42,371,516,40],"34898341","tta",[518,513,42,247,519,40],"34898407","SalimoS",[521,522,40,208,523,40],"34902488",31,"adversaryIdiot",[525,526,40,208,527,40],"34931481",32,"Existenceblinks",[529,530,42,242,531,40],"34898398",33,"bilekas",[174,533,45,310,175,40],34,[535,533,40,208,536,40],"34900736","cactusplant7374",[538,539,33,492,540,40],"34901356",35,"smsm42",[542,16,33,492,543,40],"34899124","helsontaveras18",[545,546,33,492,547,40],"34900744",37,"AndrewKemendo",[549,32,33,193,550,40],"34900899","hbn",[552,553,42,410,554,40],"34902448",39,"m3kw9",[556,557,45,332,558,40],"34915371",40,"sorokod",[560,561,42,258,562,40],"34901036",41,"tomr75",[564,561,33,492,565,40],"34901792","all2",[567,568,40,208,569,40],"34902035",42,"blueicelake121",[571,568,40,208,572,40],"34902985","gus_massa",[574,575,40,208,576,40],"34898371",43,"pknerd",[],[579,584,588,593,597,602],{"id":580,"name":581,"points":582,"category":583},"34869244","PyGWalker",712,"dev_tools",{"id":585,"name":586,"points":587,"category":583},"34897645","highlight.io",408,{"id":589,"name":590,"points":591,"category":592},"34826085","Basement Community",373,"consumer_social",{"id":594,"name":595,"points":596,"category":583},"34848583","Noya",285,{"id":598,"name":599,"points":600,"category":601},"34881286","Yobulk",233,"data",{"id":603,"name":604,"points":605,"category":606},"34901481","Jotte",212,"creative",false,"v2",{"id":610,"name":5,"line":6,"ai":611,"cat":7,"year":13,"points":14,"comments":15,"rec":28,"s":612,"top":18,"kw":613,"maker":11,"domain":10,"tier":19,"attempts":40,"strip":614,"url":9,"mix":615,"praise":87,"objection":106,"sub":20},34897773,true,-0.12,[23,24,25,26,27],"1000021300332334100314352113313314333141",[31,32,33,34],1793991060818]