15:[[47691761,{"by":"agent_anuj","descendants":2,"id":47691761,"kids":[47691769,47691802],"score":3,"time":1775662974,"title":"Is this the end for data analysts? AI agent, 500GB, 100 rounds, 134.9s, $1.66","type":"story","url":"https://medium.com/generative-ai/i-talked-to-500gb-of-retail-data-with-zero-domain-knowledge-ai-designed-a-strategy-i-never-expected-11e97b37ad04"}],[47614813,{"by":"agent_anuj","descendants":1,"id":47614813,"kids":[47614839],"score":1,"time":1775139334,"title":"Show HN: I Talked to 500GB of Retail Data with Zero Domain Knowledge","type":"story","url":"https://medium.com/generative-ai/i-talked-to-500gb-of-retail-data-with-zero-domain-knowledge-ai-designed-a-strategy-i-never-expected-11e97b37ad04"}],[47589145,{"by":"mpweiher","descendants":1,"id":47589145,"kids":[47589368],"score":2,"time":1774972033,"title":"The Conversation Is the Commit","type":"story","url":"https://aicoding.leaflet.pub/3mhxvpam4z22z"}],[47567969,{"by":"mthwsjc_","descendants":191,"id":47567969,"kids":[47571748,47569152,47569446,47568972,47568842,47568804,47569409,47571666,47569115,47568995,47569663,47577026,47568893,47570543,47568726,47576442,47568847,47570091,47571669,47568786,47576722,47569672,47569143,47568763,47570641,47569153,47569123,47569847,47569856,47568805,47569000,47568857,47568695,47569578,47569474,47569607,47569021,47568767,47570181,47570470,47569561,47571323,47568912,47575409,47571262,47569058,47591767,47570928,47568517,47568896,47568590,47570065,47571716,47568835,47568793,47568911,47568560,47568574],"score":251,"time":1774822506,"title":"Claude Code runs Git reset –hard origin/main against project repo every 10 mins","type":"story","url":"https://github.com/anthropics/claude-code/issues/40710"}],[47494696,{"by":"noemit","descendants":22,"id":47494696,"kids":[47498409,47496895,47569769,47540979,47518087,47561130,47540200,47499371,47499342,47499000,47548754,47508348,47532964,47526878,47508063,47498203,47500805,47522625,47497365,47498274],"score":41,"text":"I ran an experiment to see if CLI actually was the most intuitive format for tool calling. (As claimed by a ex-Manus AI Backend Engineer) I gave my model random scenarios and a single tool "run" - i told it that it worked like a CLI. I told it to guess commands.

it guessed great commands, but it formatted it always with a colon up front, like \n:help\n:browser\n:search\n:curl

It was trained on how terminals look, not what you actually type (you don't type the ":")

I have since updated my code in my agent tool to stop fighting against this intuition.

LLMs they learn what commands look like in documentation/artifacts, not what the human actually typed on the keyboard.

Seems so obvious. This is why you have to test your LLM and see how it naturally works, so you don't have to fight it with your system prompt.

This is Kimi K2.5 Btw.","time":1774297908,"title":"LLMs learn what programmers create, not how programmers work","type":"story"}],[47567183,{"by":"gnabgib","descendants":33,"id":47567183,"kids":[47568239,47567259,47569565,47570431,47569229,47570343,47582896,47568502,47568509,47567569],"score":56,"time":1774817515,"title":"AI isn't killing jobs, it's 'unbundling' them into lower-paid chunks","type":"story","url":"https://www.theregister.com/2026/03/24/ai_job_unbundling/"}],[47554773,{"by":"oldfrenchfries","descendants":617,"id":47554773,"kids":[47556125,47556522,47555768,47555527,47555285,47555140,47555632,47554981,47555406,47555572,47556026,47555379,47561891,47556197,47561133,47555453,47559337,47556280,47561211,47560249,47558560,47563097,47555438,47562195,47561310,47557234,47556855,47569527,47560377,47565282,47555167,47556463,47560308,47557293,47556992,47562163,47557442,47561731,47556094,47555233,47567291,47556109,47555304,47555855,47563820,47564692,47556318,47556476,47555675,47554776,47562821,47563230,47555454,47558130,47555143,47556974,47560683,47557774,47558137,47555086,47555713,47562684,47556686,47556774,47560267,47557449,47556023,47555824,47559332,47558652,47556417,47556718,47555383,47559249,47555423,47563267,47561912,47555629,47558701,47556372,47555670,47556439,47559089,47555942,47565044,47555663,47555452,47560473,47556893,47556434,47563013,47555869,47561748,47559397,47562844,47556397,47555916,47555854,47555139,47556351,47559167,47562494,47654210,47712383,47558245,47558520,47559179,47561810,47611536,47556672,47567989,47556542,47557807,47555889,47563518,47558587,47556910,47635377,47562101,47560556,47557373,47559380,47561400,47563737,47555886,47558880,47559544,47558981,47555890,47557143,47556449,47561763,47555199,47558566,47596417,47555945,47561187,47556195,47555146,47561922,47555323,47555125,47555108],"score":792,"text":"https://arxiv.org/abs/2602.14270

https://www.science.org/doi/10.1126/science.aec8352","time":1774706913,"title":"AI overly affirms users asking for personal advice","type":"story","url":"https://news.stanford.edu/stories/2026/03/ai-advice-sycophantic-models-research"}],[47564620,{"by":"jeeybee","descendants":7,"id":47564620,"kids":[47564630,47577613,47590911,47573527,47568926,47564882,47579393,47565366],"score":13,"time":1774801836,"title":"Show HN: Pglens – 27 read-only PostgreSQL tools for AI agents via MCP","type":"story","url":"https://github.com/janbjorge/pglens"}]]

user:	agent_anuj
created:	March 29, 2026
karma:	5
about:	Agentic AI Architect with 20 years of enterprise platform experience.

1.	Is this the end for data analysts? AI agent, 500GB, 100 rounds, 134.9s, $1.66(medium.com) 3 points by agent_anuj 39 days ago \| 2 comments
2.	Show HN: I Talked to 500GB of Retail Data with Zero Domain Knowledge(medium.com) 1 points by agent_anuj 46 days ago \| 1 comment