raducu's favorites | Hacker News

1.		AI agents are starting to eat SaaS (martinalderson.com)
		137 points by jnord 11 hours ago \| 158 comments
2.		OpenAI are quietly adopting skills, now available in ChatGPT and Codex CLI (simonwillison.net)
		569 points by simonw 2 days ago \| 314 comments
3.		Is it a bubble? (oaktreecapital.com)
		333 points by saigrandhi 4 days ago \| 564 comments
4.		The universal weight subspace hypothesis (arxiv.org)
		358 points by lukeplato 6 days ago \| 132 comments
5.		LLM from scratch, part 28 – training a base model from scratch on an RTX 3090 (gilesthomas.com)
		539 points by gpjt 12 days ago \| 120 comments
6.		Google Titans architecture, helping AI have long-term memory (research.google)
		586 points by Alifatisk 7 days ago \| 186 comments
7.		Nested Learning: A new ML paradigm for continual learning (research.google)
		152 points by themgt 7 days ago \| 10 comments
8.		How elites could shape mass preferences as AI reduces persuasion costs (arxiv.org)
		703 points by 50kIters 11 days ago \| 660 comments
9.		How to Synthesize a House Loop (loopmaster.xyz)
		259 points by stagas 17 days ago \| 100 comments
10.		Mistral 3 family of models released (mistral.ai)
		826 points by pember 12 days ago \| 236 comments
11.		Codex, Opus, Gemini try to build Counter Strike (instantdb.com)
		287 points by stopachka 16 days ago \| 132 comments
12.		AI agents find $4.6M in blockchain smart contract exploits (anthropic.com)
		215 points by bpierre 13 days ago \| 121 comments
13.		DeepSeek-v3.2: Pushing the frontier of open large language models [pdf] (huggingface.co)
		982 points by pretext 13 days ago \| 465 comments
14.		All it takes is for one to work out (alearningaday.blog)
		813 points by herbertl 15 days ago \| 414 comments
15.		Statistical Process Control in Python (timothyfraser.com)
		230 points by lifeisstillgood 19 days ago \| 73 comments
16.		CS234: Reinforcement Learning Winter 2025 (stanford.edu)
		207 points by jonbaer 19 days ago \| 60 comments
17.		Claude Opus 4.5 (anthropic.com)
		1113 points by adocomplete 20 days ago \| 506 comments
18.		Adversarial poetry as a universal single-turn jailbreak mechanism in LLMs (arxiv.org)
		384 points by capgre 24 days ago \| 189 comments
19.		210 IQ Is Not Enough (taylor.town)
		177 points by surprisetalk 24 days ago \| 279 comments
20.		AI is a front for consolidation of resources and power (chrbutler.com)
		545 points by delaugust 25 days ago \| 448 comments
21.		Gemini 3 (blog.google)
		1735 points by preek 26 days ago \| 1056 comments
22.		Messing with scraper bots (herman.bearblog.dev)
		245 points by HermanMartinus 30 days ago \| 84 comments
23.		Spec-Driven Development: The Waterfall Strikes Back (marmelab.com)
		225 points by vinhnx 30 days ago \| 191 comments
24.		Dark Pattern Games (darkpattern.games)
		350 points by robotnikman 28 days ago \| 138 comments
25.		Neuroscientists track the neural activity underlying an “aha” (quantamagazine.org)
		164 points by wjb3 28 days ago \| 54 comments
26.		PgFirstAid: PostgreSQL function for improving stability and performance (github.com/randoneering)
		117 points by yakshaving_jgt 28 days ago \| 23 comments
27.		Maybe you’re not trying (usefulfictions.substack.com)
		448 points by eatitraw 29 days ago \| 188 comments
28.		Honda: 2 years of ml vs 1 month of prompting - heres what we learned (levs.fyi)
		320 points by Ostatnigrosh 34 days ago \| 108 comments
29.		A race condition in Aurora RDS (hightouch.com)
		241 points by theanomaly 30 days ago \| 82 comments
30.		Reverse Engineering Yaesu FT-70D Firmware Encryption (landaire.net)
		162 points by austinallegro 32 days ago \| 23 comments
		More