18:[[47111440,{"by":"jakozaur","descendants":98,"id":47111440,"kids":[47112509,47111925,47115255,47112403,47113023,47111835,47111486,47113130,47112156,47113415,47112208,47134040,47111617,47112231,47112203,47114549,47112079,47114582,47113725,47112490,47114645,47113651,47113433,47112582,47113256,47112548,47115025,47112649,47111711,47112909,47112917,47114129],"score":245,"time":1771771844,"title":"We hid backdoors in ~40MB binaries and asked AI + Ghidra to find them","type":"story","url":"https://quesma.com/blog/introducing-binaryaudit/"}],[47086181,{"by":"sidnarsipur","descendants":455,"id":47086181,"kids":[47086634,47090825,47087147,47087187,47087367,47086555,47102601,47096893,47086532,47092371,47087021,47094620,47088168,47087427,47086855,47086583,47087717,47086564,47091143,47086948,47093999,47086576,47089386,47086874,47089202,47096517,47087440,47090153,47088532,47101784,47086639,47087347,47087491,47086797,47100912,47087620,47089825,47090835,47088082,47091842,47090715,47103781,47094073,47086712,47088298,47094176,47093465,47086826,47096575,47094825,47091066,47087512,47086549,47096346,47089168,47095328,47087409,47086706,47103568,47103573,47088102,47086469,47094251,47086903,47099551,47087050,47086814,47091261,47087454,47086949,47086632,47089128,47091946,47086539,47086523,47086641,47091932,47087480,47091768,47091311,47086653,47090989,47092916,47096660,47087258,47086878,47087664,47086587,47087188,47093010,47091236,47087866,47091817,47091572,47086927,47086851,47089079,47086862,47094991,47086694,47086573,47087082,47088072,47087016,47091573,47097400,47087088,47088654,47087080,47101587,47094150,47097141,47087355,47087807,47087382,47086676,47097763,47086455,47088285,47094949,47091404,47090965,47089561,47090196,47099200,47087858,47087547,47098543,47103604,47086880,47086930,47086817,47086614],"score":847,"time":1771583572,"title":"The path to ubiquitous AI (17k tokens/sec)","type":"story","url":"https://taalas.com/the-path-to-ubiquitous-ai/"}],[47075400,{"by":"domysee","descendants":95,"id":47075400,"kids":[47075737,47075761,47075720,47075697,47075660,47076047,47075930,47087870,47076017,47075699,47078816,47075891,47076127,47075965,47076585,47079690,47080837,47075972,47076054,47081081,47076070,47092243,47076343,47076836,47076787,47075837,47076321,47076145,47075769,47076607,47075826,47102968,47075724,47076140,47080063,47094281,47075604,47076007,47076436,47076006],"score":105,"time":1771518026,"title":"AI made coding more enjoyable","type":"story","url":"https://weberdominik.com/blog/ai-coding-enjoyable/"}],[47074735,{"by":"MallocVoidstar","descendants":914,"id":47074735,"kids":[47076453,47079966,47080308,47075566,47076937,47076484,47076733,47075747,47075916,47075750,47076703,47075340,47075498,47075664,47076285,47075386,47075618,47085303,47082391,47076344,47082908,47078572,47080142,47077551,47075581,47076088,47082937,47077009,47078265,47079104,47076172,47075948,47078302,47075751,47076500,47076958,47081311,47075355,47085311,47081653,47076019,47105644,47086857,47083519,47075600,47075439,47075961,47086593,47075681,47085630,47075954,47075669,47075529,47087015,47076086,47076511,47082960,47095222,47075923,47076486,47098498,47077070,47077606,47076036,47075367,47081106,47076810,47077238,47075939,47080637,47075881,47082350,47076328,47084909,47086800,47087555,47096362,47075337,47076049,47075825,47086972,47075599,47075725,47080818,47080993,47075755,47077934,47080030,47077207,47075586,47081745,47084282,47080173,47084686,47083092,47075579,47078141,47085015,47081554,47084863,47076133,47079826,47076504,47075798,47129163,47076181,47075535,47079418,47079204,47076432,47080386,47077402,47075594,47078599,47075304,47084429,47087035,47081414,47075740,47075946,47075728,47076934,47078180,47079354,47077693,47080146,47075689,47076863,47077328,47077507,47076572,47077877,47078959,47077814,47077346,47081090,47083988,47076433,47075223,47077912,47076031,47076074],"score":963,"text":"Preview: https://console.cloud.google.com/vertex-ai/publishers/google...

Card: https://deepmind.google/models/model-cards/gemini-3-1-pro/","time":1771514397,"title":"Gemini 3.1 Pro","type":"story","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/"}],[47050488,{"by":"adocomplete","descendants":1226,"id":47050488,"kids":[47053425,47056847,47054498,47059997,47051089,47053819,47050837,47054292,47051472,47054665,47051221,47052194,47052072,47056885,47055281,47051172,47053713,47055868,47050819,47051286,47051412,47051134,47055729,47051407,47051233,47050969,47050839,47052692,47051076,47050863,47051492,47050751,47056550,47051497,47116599,47051544,47106989,47054547,47055179,47051691,47060645,47060616,47052418,47050805,47052984,47052703,47060955,47058120,47052978,47055378,47057006,47050814,47050960,47065612,47051804,47051120,47055910,47051277,47051250,47051335,47052942,47051397,47059409,47051446,47055213,47055178,47055148,47059314,47051706,47059012,47062111,47055871,47050833,47052278,47054815,47051627,47061065,47060762,47059537,47059549,47055690,47056978,47051174,47050890,47057026,47050934,47053903,47050746,47051119,47052512,47050711,47057165,47053986],"score":1346,"text":"https://www.anthropic.com/claude-sonnet-4-6-system-card [pdf]

https://x.com/claudeai/status/2023817132581208353 [video]","time":1771350532,"title":"Claude Sonnet 4.6","type":"story","url":"https://www.anthropic.com/news/claude-sonnet-4-6"}],[47049227,{"by":"GregorStocks","descendants":83,"id":47049227,"kids":[47051666,47050619,47051033,47050559,47050946,47054647,47053825,47057759,47050560,47050869,47050521,47051194,47051401,47052596,47051090,47054366,47051795,47050653,47050474,47051305,47055594,47050573,47050487,47055232,47049228],"score":117,"text":"I've been teaching LLMs to play Magic: The Gathering recently, via MCP tools hooked up to the open-source XMage codebase. It's still pretty buggy and I think there's significant room for existing models to get better at it via tooling improvements, but it pretty much works today. The ratings for expensive frontier models are artificially low right now because I've been focusing on cheaper models until I work out the bugs, so they don't have a lot of games in the system.","time":1771345369,"title":"Show HN: I taught LLMs to play Magic: The Gathering against each other","type":"story","url":"https://mage-bench.com/"}],[46905502,{"by":"mbh159","descendants":14,"id":46905502,"kids":[46906342,46905503,46908259,46906402,46906456],"score":10,"time":1770326409,"title":"Live agent face-off in CivBench: Claude Opus 4.6 vs. GPT-5.2","type":"story","url":"https://www.clashai.live"}]]

1.	Show HN: CivBench a long-horizon AI benchmark for multi-agent games(clashai.live) 12 points by mbh159 81 days ago \| 24 comments
2.	Live agent face-off in CivBench: Claude Opus 4.6 vs. GPT-5.2(clashai.live) 10 points by mbh159 101 days ago \| 14 comments

user:	mbh159
created:	September 25, 2019
karma:	13