SIGNAL

Tuesday, July 21, 2026

17 stories · 5 min read

★ Must ReadAnthropic’s landmark $1.5B copyright settlement is approved

The final approval settles one case, but it doesn't resolve the broader issue of using copyrighted works to train AI models.

TechCrunch AI· Kirsten Korosec· 2 min· 12h agoSource →

01QUICK HITS

01

China’s open-weights AI strategy is winning

Trending on Hacker News with 993 points and 801 comments.

Hacker News · 1 min

→02

Airport Simulator

Trending on Hacker News with 736 points and 142 comments.

Hacker News · 1 min

→03

Claude Fable produced a counterexample to the Jacobian Conjecture

Trending on Hacker News with 693 points and 447 comments.

Hacker News · 1 min

→04

Firefox 153 available with support for Vulkan video decoding, JPEG-XL

Trending on Hacker News with 242 points and 55 comments.

Hacker News · 1 min

→05

Google is working on a new AI chip designed to make Gemini more efficient

Alphabet, Google's parent company, is reportedly working on a new chip designed to make its Gemini models run much more efficiently.

TechCrunch AI · 2 min

→

03THE TENSION2▸

PERSPECTIVE A

“China has all but caught up. The US is not going to “win” the AI war. Here’s what we should do instead.”

— Gary Marcus

Gary Marcus — China has all but caught up. The US is not going to “win” th...

Source →

♦vs♦

PERSPECTIVE B

“Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling”

— Hacker News

Trending on Hacker News with 302 points and 298 comments.

Source →

04RESEARCH3▸

★ Must ReadWho's afraid of Chinese models?

Trending on Hacker News with 328 points and 212 comments.

Hacker News· 1 min· yesterdaySource →

★ Must ReadHow we measured AI writing across arXiv, and where the measurement breaks

Trending on Hacker News with 197 points and 144 comments.

Hacker News· 1 min· 20h agoSource →

Import AI 465: Open vs closed gaps; Kimi K3; Demis' big policy plan

The singularity will be seen in hindsight as an interregnum

Import AI· Jack Clark· 3 min· yesterdaySource →

05BUILDERS2▸

Connect Your Tools to AI 🔗

How and why to link AI to your favorite tools

Wonder Tools· Jeremy Caplan· 2 min· 2d agoSource →

The Hidden Geometry Behind AI Reasoning

The geometry behind chain-of-thought, reasoning failures, and the paths models take to an answer.

AI Made Simple· Devansh· 2 min· yesterdaySource →

06GO DEEP1▸

Agentic World Models

Creating better language agents by teaching them to model their environment...

Cameron Wolfe· Cameron R. Wolfe, Ph.D.· 3 min· yesterdaylong readSource →

07VOICES3▸

“Anthropic’s landmark $1.5B copyright settlement is approved”

— Kirsten Korosec, TechCrunch AI