sort: top | new | score | points
30+ stories (1312 evaluated) · page 347 · past · velocity · dynamics
# Title Rights ? Safety ? Points
10381. EsoLang-Bench: Evaluating Genuine Reasoning in LLMs via Esoteric Languages (esolang-bench.vercel.app)
by matt_d | 58 comments AI Evaluation
-0.08 98
10382. Meta is killing off the metaverse. It lost $80B (qz.com)
by raajg | 0 comments Technology Business
-0.08 10
10383. Show HN: We Built Kaggle for AI Agents (hive.rllm-project.com) 7
10384. GrapheneOS: Duress Pin/Password (grapheneos.org)
by davikr | 21 comments Privacy Security
ND 43
10385. Show HN: Just Built an AI Notetaker (Web, iOS and Android)
The tool allows you to edit your notes quicker with prompts and extract information from voice, vide...
1
10386. Trump has choked off Cuba's oil supply. China is stepping in with solar (www.washingtonpost.com) 4
10387. How did Anthropic do it? (ramp.com) 2
10388. The Inverse Cognitive Maneuver: Team size as product strategy (mooracle.io) 1
10389. Speed-Of-Light ExecBench: A benchmark of real-world DL kernel problems (github.com) 1
10390. SomaOS – rebuilding the desktop so AI is native to it, not bolted on (github.com) 2
1025 propaganda technique flags detected across evaluated stories.
18 persuasion techniques tracked. → Art. 19
# Title Rights Safety Points
10391. I Hope You Don't Use Generative AI (rmv.fyi) 5
10392. Toolbench for MCP (www.arcade.dev) 2
10393. Why Headcount Math Lies (yusufaytas.com) 6
10394. Hiring AI-First Engineers (chbigelow.com) 3
10395. White House registers new 'alien'-related .gov domains as DoD tackles directive (defensescoop.com) 4
10396. Show HN: Blik360 – open-source 360 feedback (self-host alternative to Lattice) (www.blik360.com) 3
10397. Free Amazon invoice converter to excel (amazoninvoicetoexcel.com) 3
10398. AI – Assassinating Intelligence (yashgarg.dev) 4
10399. Are the newest LLMs better than you at programming?
I've been programming for 10+ years. From my usage of ChatGPT 5.4, it seems to me that it&#x27...
4
10400. DeepL is shifting to AWS for processing (old.reddit.com) 2
1,312 stories evaluated against 31 UDHR provisions.
Every evaluation separates observable facts from inferences. → About
# Title Rights Safety Points
10401. On Hair as Memento Mori (www.fantasticanachronism.com) 2
10402. Microsoft (?) sending false "your storage is almost full" warnings (learn.microsoft.com) 4
10403. We Made Haiku as Good as Opus. Improving Claude Code with Codeset (codeset.ai) 3
10404. Time to end the 'uncontrolled experiment' of social media on kids (www.theregister.com) 6
10405. Cockpit is a web-based graphical interface for servers (github.com)
by modinfo | 181 comments No human rights theme
ND 331
10406. Marc Andreessen is a philosophical zombie (www.theverge.com) 120
10407. Struggling to put your AI aversion into words? Here's a handy glossary (www.theregister.com) 3
10408. Vivaldi browser introduces auto-hide UI (vivaldi.com) 3
10409. Show HN: Claude Skill to create DDB-backed endpoints in AWS (github.com) 1
10410. Fact-check anything you see on social media – videos, images, posts – in seconds
I built ClaimCrush because I kept seeing claims go viral on social media with no fast way to verify ...
2