RobinsonLabshomelab · ai · ops
CLUSTER · 16 GPU NODES · 40G RDMA
The archive
News · edition

Edition, Sun, 02 Aug 2026.

4 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

Claude Fable 5 and restricted Mythos 5 set new benchmarks, solving a 50-million-line code migration in a day and generating novel drug candidates autonomously.

1 source
Lead
this edition
Still developing
The new processor aims to optimize performance-per-watt for large language model inference, reducing reliance on Nvidia GPUs.
Apple regains crown as world's most valuable company, overtaking NvidiaInvestors reshuffled tech's top ranks as Apple's market cap hit roughly $4.88 trillion, surpassing Nvidia's $4.86 trillion on Friday.The GuardianReuters
China launches WAICO, a new global AI governance body with 29 founding nationsThe World Artificial Intelligence Cooperation Organization was officially signed on July 16 in Shanghai, positioning AI development as a shared global endeavor rather than a single-nation pursuit.The DiplomatXinhua via CPPCC
Moonshot AI's Kimi K3 becomes world's largest open-weight model, tops code benchmarksThe 2.8-trillion-parameter model released July 16 ranked first in the Frontend Code Arena with 1,679 points, beating Claude Fable 5 in blind developer testing.Tom's HardwareArena.ai
ServiceNow hits $1B AI ACV and raises full-year outlook in strong Q2 2026 earningsServiceNow reported Q2 2026 revenue of $3.987B (up 24% YoY) and announced its AI platform crossed $1B in annual contract value, prompting a raised full-year subscription revenue forecast.Outlet
CEO Mark Zuckerberg cited confidence in the investment despite market pushback on the massive capital expenditure increase.
SpaceXAI launches Grok 4.5 as an Opus-class model at lower costThe new model excels at coding, agentic tasks, and knowledge work with twice the token efficiency of competitors.SpaceXAITechCrunch
xAI releases Grok 4.5 with 1.5 trillion parameter MoE architecture and competitive pricingThe new model launched July 9 alongside GPT-5.6, featuring training on Cursor data and achieving 83.3 percent on Terminal-Bench 2.1 while using approximately 25 percent fewer output tokens than Opus 4.8.AIapps

AI & Technology

AI & Technology
Generative AI providers with over one million monthly California users must now embed provenance data and offer free detection tools starting August 2.
2 sources
AI & Technology

Anthropic debuts Mythos-class AI with record coding and science leaps

Claude Fable 5 and restricted Mythos 5 set new benchmarks, solving a 50-million-line code migration in a day and generating novel drug candidates autonomously.

Claude Fable 5 and restricted Mythos 5 set new benchmarks, solving a 50-million-line code migration in a day and generating novel drug candidates autonomously.

Anthropic released two fifth-generation models this week, marking a shift toward specialized, high-capability AI. Claude Fable 5, available generally, achieved 80.3% on the SWE-Bench Pro coding benchmark, outperforming previous Opus and GPT models, and completed a massive codebase migration for Stripe in a single day. The restricted Mythos 5, reserved for select partners, demonstrated breakthrough performance in science and cybersecurity, designing viable drug candidates without human intervention and ranking as the strongest cybersecurity model to date.

1 source
The Decoder
AI & Technology

California’s AI Transparency Act becomes operative today, mandating visible labels and free detection tools

Generative AI providers with over one million monthly California users must now embed provenance data and offer free detection tools starting August 2.

Generative AI providers with over one million monthly California users must now embed provenance data and offer free detection tools starting August 2.

The California AI Transparency Act (SB 942), amended by AB 853, officially went into effect today, targeting large generative AI providers accessible to California users. Covered providers must now offer a free AI detection tool, embed machine-readable provenance metadata in images, video, and audio, and provide a visible label option for AI-generated content. Non-compliance carries a penalty of $5,000 per day per violation, with the law applying specifically to media outputs rather than text.

2 sources
Agora Intelligence
AuditSocials
AI & Technology

Meta raises 2026 AI infrastructure spending floor to $130 billion as stock dips on ROI questions

CEO Mark Zuckerberg cited confidence in the investment despite market pushback on the massive capital expenditure increase.

CEO Mark Zuckerberg cited confidence in the investment despite market pushback on the massive capital expenditure increase.

Meta Platforms increased its capital expenditure forecast for 2026 to a range of $130 billion to $145 billion, driven by aggressive data center and custom silicon development. The announcement followed a Q1 earnings report showing a 33% revenue jump, yet the stock fell over 6% in after-hours trading as investors questioned the timeline for return on investment. Zuckerberg noted the company is building custom chips with Broadcom and AMD, aiming for a "first-principles understanding" of content in its AI models.

1 source
Fortune
AI & Technology

OpenAI unveils Jalapeño, its first custom inference chip designed with Broadcom

The new processor aims to optimize performance-per-watt for large language model inference, reducing reliance on Nvidia GPUs.

The new processor aims to optimize performance-per-watt for large language model inference, reducing reliance on Nvidia GPUs.

OpenAI introduced Jalapeño, a custom silicon chip co-developed with Broadcom specifically for running its large language model inference workloads. Early tests indicate significantly better performance-per-watt compared to current state-of-the-art alternatives, targeting cost efficiency in real-time coding and chat applications. This move mirrors similar strategies by Google and Amazon, allowing OpenAI to optimize the full stack from model to hardware, with deployment expected in late 2026.

1 source
TechCrunch
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.

Back to the archive Latest edition