anukin's favorites | Hacker News

1.		Vibe Code Warning – A personal casestudy (github.com/jackdoe)
		393 points by jackdoe 2 days ago \| 309 comments
2.		How cops can get your private online data (eff.org)
		295 points by jamesgill 2 days ago \| 64 comments
3.		LLMs are steroids for your Dunning-Kruger (bytesauna.com)
		388 points by gridentio 2 days ago \| 298 comments
4.		Study finds memory decline surge in young people (onepercentrule.substack.com)
		200 points by drcwpl 4 days ago \| 194 comments
5.		Study identifies weaknesses in how AI systems are evaluated (ox.ac.uk)
		414 points by pseudolus 4 days ago \| 192 comments
6.		Becoming a compiler engineer (rona.substack.com)
		293 points by lalitkale 5 days ago \| 151 comments
7.		Angel Investors, a Field Guide (jeanyang.com)
		200 points by azhenley 5 days ago \| 55 comments
8.		Cerebras Code now supports GLM 4.6 at 1000 tokens/sec (cerebras.ai)
		194 points by nathabonfim59 5 days ago \| 128 comments
9.		Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model (moonshotai.github.io)
		933 points by nekofneko 6 days ago \| 425 comments
10.		Show HN: I scraped 3B Goodreads reviews to train a better recommendation model (book.sv)
		603 points by costco 7 days ago \| 259 comments
11.		Leaving Meta and PyTorch (soumith.ch)
		720 points by saikatsg 6 days ago \| 172 comments
12.		Why aren't smart people happier? (theseedsofscience.pub)
		525 points by zdw 7 days ago \| 622 comments
13.		The trust collapse: Infinite AI content is awful (arnon.dk)
		236 points by arnon 6 days ago \| 211 comments
14.		NY school phone ban has made lunch loud again (gothamist.com)
		433 points by hrldcpr 7 days ago \| 334 comments
15.		The Case That A.I. Is Thinking (newyorker.com)
		278 points by ascertain 9 days ago \| 1009 comments
16.		Meta readies $25B bond sale as soaring AI costs trigger stock sell-off (ft.com)
		120 points by 1vuio0pswjnm7 11 days ago \| 188 comments
17.		Facts about throwing good parties (atvbt.com)
		962 points by cjbarber 10 days ago \| 412 comments
18.		Autodesk's John Walker Explained HP and IBM in 1991 (2015) (cringely.com)
		136 points by suioir 14 days ago \| 71 comments
19.		Context engineering (chrisloy.dev)
		98 points by chrisloy 10 days ago \| 65 comments
20.		The Smol Training Playbook: The Secrets to Building World-Class LLMs (huggingface.co)
		265 points by kashifr 13 days ago \| 19 comments
21.		Backpropagation is a leaky abstraction (2016) (karpathy.medium.com)
		352 points by swatson741 11 days ago \| 159 comments
22.		Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch (tongyi-agent.github.io)
		365 points by meander_water 10 days ago \| 153 comments
23.		Show HN: Why write code if the LLM can just do the thing? (web app experiment) (github.com/samrolken)
		435 points by samrolken 11 days ago \| 323 comments
24.		Claude Code can debug low-level cryptography (filippo.io)
		471 points by Bogdanp 11 days ago \| 208 comments
25.		How I use every Claude Code feature (sshh.io)
		532 points by sshh12 11 days ago \| 188 comments
26.		You can't refuse to be scanned by ICE's facial recognition app, DHS document say (404media.co)
		601 points by nh43215rgb 11 days ago \| 509 comments
27.		Signs of introspection in large language models (anthropic.com)
		183 points by themgt 13 days ago \| 126 comments
28.		The profitable startup (linear.app)
		248 points by doppp 12 days ago \| 89 comments
29.		Ask HN: Who uses open LLMs and coding assistants locally? Share setup and laptop
		350 points by threeturn 12 days ago \| 192 comments
30.		How to build silos and decrease collaboration on purpose (rubick.com)
		144 points by gpi 12 days ago \| 54 comments
		More