Weekly Best
2791.
How Much of AI Labs' Research Is Safety? (fi-le.net)
2792.
The End (focalcurve.com)
2793.
How to Kill a Tech Monopoly (fourplex.net)
2794.
Show HN: Zitrus – A 3DS SDK written in pure Zig (codeberg.org)
2795.
In a world of AI text, speech still reigns supreme (techxplore.com)
2796.
Tapes: transparent AI agent telemetry (johncodes.com)
2797.
The Video Software That Powers the Internet (and Almost Nobody Knows It Exists) (roughcut.heyeddie.ai)
2798.
How to Do AI-Assisted Engineering (newsletter.eng-leadership.com)
2799.
American Diner Gothic (thenewatlantis.com)
2800.
Codex Subagents (developers.openai.com)
2801.
Show HN: Solution for Prompt Injection of AI Agents (github.com)
2802.
Show HN: AgentDrive – Persistent file storage for AI agents (getagentdrive.com)
2803.
Quick Comments on the SCOTUS Cox vs. Sony Ruling (blog.ericgoldman.org)
2804.
I stopped promoting my app for 2 days and traffic instantly dropped (superbrain.base44.app)
2805.
Hyperagents (arxiv.org)
2806.
The First Video Game Was Just a Box in the Corner of a Bar (lithub.com)
2807.
Show HN: My Hyperliquid Trading Terminal (aulico.com)
2808.
Show HN: Behavioral analysis of your career decisions from your LinkedIn (coached.com)
2809.
Reasoning-Based Personalized Generation for Users with Sparse Data (arxiv.org)
2810.
Linux Kodachi 9 (kodachi.cloud)
2811.
Why colleges are turning to oral exams to combat AI (apnews.com)
2812.
Infinite Jest Extraction (mattlakeman.org)
2813.
RustSec Integrity Breach Hides Dangerous Crypto Flaw (flyingpenguin.com)
2814.
Stop Embedding Your Corpus Blindly (decompressed.io)
2815.
Show HN: Threat hunting command system for agentic IDEs (github.com)
2816.
Prompt Repetition Improves Non-Reasoning LLMs (arxiv.org)
2817.
Stephen Wolfram and Matt Mullenweg Talk About AI [video] (youtube.com)
2818.
Show HN: RoverBook – PostHog for AI Agents (github.com)
2819.
Show HN: Dbt-skillz compiles your dbt project into a Claude Code skill (github.com)
2820.
Lessons from Pyre That Shaped Pyrefly – Pyrefly (pyrefly.org)