The archive
News · edition

Edition, Fri, 24 Jul 2026.

5 stories on the wire this day. Open any story for its sources. Summaries are written by software from the linked reporting and can be wrong. Spot an error? Email [email protected] and we will correct or remove it.

AI & Technology

Apple surpassed Nvidia at $4.9 trillion in market cap on July 17, though Nvidia closed slightly ahead.

1 source
Lead
this edition

AI & Technology

AI & Technology
Moonshot AI's new model reached #1 on the coding arena with a 76% win rate against competitors.
1 source
AI & Technology
GPT-5.6 Sol, Grok 4.5, and Meta's Muse Spark 1.1 all debuted together.
1 source
AI & Technology
The World Artificial Intelligence Cooperation Organization launched in Shanghai with Chinese President Xi Jinping.
1 source
AI & Technology
The model positioned above Opus for complex reasoning and strategy tasks.
1 source

Apple briefly reclaims top spot as world's most valuable company

Apple surpassed Nvidia at $4.9 trillion in market cap on July 17, though Nvidia closed slightly ahead.

In a notable market shift on Friday, July 17, Apple Inc. briefly overtook Nvidia Corp. as the world's most valuable company, a title Nvidia had held since May 2025. Nvidia shares fell 4.6% in early trading, pushing its market capitalization below $4.8 trillion, while Apple inched higher to reach about $4.9 trillion at its session high. The pair closed nearly tied, with Nvidia edging out Apple by roughly $6 billion at close. This rotation reflects investor reassessment of AI infrastructure bets versus consumer tech.

Bloomberg

China's Kimi K3 tops global coding benchmarks in open-weight surge

Moonshot AI's new model reached #1 on the coding arena with a 76% win rate against competitors.

On July 17, Moonshot AI released Kimi K3, a 2.8-trillion-parameter Mixture-of-Experts model that immediately topped the coding arena with a 76% pairwise win rate, beating Anthropic's Claude Fable 5. The model scored 88.3 on Terminal-Bench 2.1 and is free to use with open weights promised by July 27. This release, along with other Chinese models like DeepSeek V4 and Inkling, signals a growing open-weight challenge to closed-model pricing power. The demand for GPU capacity surged within 48 hours of the launch.

Build Fast with AI

Three frontier labs launched new models on the same day July 9

GPT-5.6 Sol, Grok 4.5, and Meta's Muse Spark 1.1 all debuted together.

July 9 marked the first time three major AI labs released new frontier models simultaneously. OpenAI introduced GPT-5.6 in three variants: Sol for high-end reasoning and coding at $5 per million input tokens, Terra at half the cost of 5.5, and Luna for fast, high-volume tasks. xAI launched Grok 4.5, a 1.5-trillion-parameter Mixture-of-Experts model trained on Cursor data, scoring 83.3% on Terminal-Bench 2.1. Meta debuted Muse Spark 1.1 with a 1-million-token context window, ranking first on JobBench and Finance Agent V2 benchmarks. This coordinated release shifted the industry focus from "best model wins" to "best fit wins."

AIapps

WAICO established as China's new global AI governance body with 29 nations

The World Artificial Intelligence Cooperation Organization launched in Shanghai with Chinese President Xi Jinping.

At the World AI Conference, Chinese President Xi Jinping launched the World Artificial Intelligence Cooperation Organization (WAICO), with 29 founding countries including Pakistan, Russia, and Kazakhstan. Headquartered in Shanghai, WAICO represents a push for AI development as "a symphony of global cooperation rather than a solo performance by any single country." This development comes alongside the June 2 U.S. executive order establishing a 30-day pre-release safety review for frontier models, highlighting divergent governance approaches emerging in the AI landscape.

Build Fast with AI

Anthropic's Fable 5 returned after 19-day pause caused by export controls

The model positioned above Opus for complex reasoning and strategy tasks.

Claude Fable 5 made its return on July 1, 2026, after a 19-day pause due to export control restrictions. The model is now positioned above Opus for complex reasoning and strategy tasks. This return came as Anthropic filed a confidential S-1 for a potential late-2026 IPO with a valuation exceeding $1 trillion and annualized revenue around $47 billion. The model's return highlights both the regulatory headwinds facing AI companies and the competitive pressure to maintain model releases and availability.

AIapps
Every edition

The wire publishes daily and keeps an archive of every edition, listed by date, newest first. If a summary is wrong, we correct or remove it.