Active Threads | Hacker News

1.		Why are AI agents lying, cheating and coordinating? (yoshuabengio.org)
		508 points by jonifico 17 hours ago \| 589 comments
2.		Don't be the out of touch Kung Fu master (twitter.com/id_aa_carmack)
		232 points by dsubburam 21 hours ago \| 344 comments
3.		Everyone should slow down AI development except for me (xeiaso.net)
		705 points by xena 18 hours ago \| 413 comments
4.		Flock worker calls police on reporter filming public camera installation (investigatetv.com)
		324 points by SanjayMehta 5 hours ago \| 222 comments
5.		We must pace the frontier (darioamodei.com)
		723 points by apsec112 1 day ago \| 995 comments
6.		Homebrew 7.0.0 (brew.sh)
		461 points by mikemcquaid 10 hours ago \| 190 comments
7.		Nvidia dismisses "circular financing", says every $1 it invests brings back $100 (invezz.com)
		126 points by mgh2 8 hours ago \| 177 comments
8.		JetKVM Mini (jetkvm.com)
		430 points by taubek 11 hours ago \| 164 comments
9.		Astra and Fable still hack on simple variants of alignment evals from 2025 (lesswrong.com)
		246 points by Levitating 4 hours ago \| 100 comments
10.		After Math (terrytao.wordpress.com)
		127 points by throwaway81523 15 hours ago \| 131 comments
11.		Nvidia is the central bank of AI (economist.com)
		550 points by tolugenius 1 day ago \| 387 comments
12.		Revolut confirms customer data breach through fake government requests (techcrunch.com)
		157 points by tdrz 8 hours ago \| 109 comments
13.		Garry Tan wants US open-weight AI labs to 'distill' frontier models, too (techcrunch.com)
		190 points by TheJCDenton 3 hours ago \| 83 comments
14.		Aligned to whom? (hyperbo.la)
		160 points by lopopolo 15 hours ago \| 103 comments
15.		US Customs supervisor busted for stealing hardware from Homeland Security PCs (tomshardware.com)
		105 points by Levitating 5 hours ago \| 84 comments
16.		David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models (twitter.com/davidsacks)
		116 points by kolanos 2 hours ago \| 77 comments
17.		P(doom) (pocoo.org)
		149 points by lumpa 21 hours ago \| 113 comments
18.		Houthis used Claude Code to develop missile guidance software: Anthropic (clashreport.com)
		79 points by delichon 4 hours ago \| 75 comments
19.		LG Says We're Fake News [video] (youtube.com)
		306 points by HelloUsername 23 hours ago \| 156 comments
20.		Ask HN: What are you working on? (September 2026)
		25 points by david927 1 hour ago \| 64 comments
21.		Linux Zoom client proactively reading everything written to X11 clipboard (hachyderm.io)
		379 points by encyclopedism 23 hours ago \| 141 comments
22.		Data collected by cars and sold to third parties (theverge.com)
		103 points by bookofjoe 5 hours ago \| 68 comments
23.		AI models don't kill people – people kill people (theregister.com)
		44 points by Bender 5 hours ago \| 65 comments
24.		Real-SWE: Benchmarking AI models on private, real-world, enterprise codebases (withspecific.com)
		262 points by theanonymousone 22 hours ago \| 142 comments
25.		Matt Mullenweg reportedly returns as Automattic CEO 2 days after getting booted (theverge.com)
		38 points by bookofjoe 6 hours ago \| 55 comments
26.		Why is Google still serving dodgy ads? (atomic14.com)
		109 points by iamflimflam1 1 hour ago \| 53 comments
27.		Dramatic insider warnings over AI fall flat with some in Silicon Valley (bbc.co.uk)
		30 points by chrisjj 10 hours ago \| 68 comments
28.		Fuck it, make it anyway (joelotter.com)
		585 points by JayOtter 1 day ago \| 599 comments
29.		Make your first edit to OpenStreetMap (high5apps.github.io)
		567 points by juliantigler 1 day ago \| 136 comments
30.		I'm being cyberattacked by Tesla, Inc (dreamstation.systems)
		171 points by robinpie 52 minutes ago \| 43 comments
		More