Sakhanda Wire
NVDA $155.46 -0.44% MSFT $506.24 -2.27% GOOGL $280.20 -1.24% META $618.87 -1.20% AMZN $257.89 +0.05%

Latest news

Older stories

TechCrunch

Crusoe sells unfinished Wyoming data center project to Google

Crusoe, a Denver-based data center developer valued at $30.9 billion, has sold an unfinished project in southeastern Wyoming to Google. The eight-year-old company confirmed the deal, while Google said it plans to invest $17 billion to develop the site. The purchase price was not disclosed.

The Verge

China reportedly recruited executive to gather intelligence on Jensen Huang

China’s Ministry of State Security reportedly recruited Megaspeed executive Alice Huang to gather information from Nvidia CEO Jensen Huang, according to a report cited in the article. The agency was said to be seeking details about American AI chips and possibly trying to influence Nvidia to make its most advanced chips available in China.

The Verge

AI data center expansion faces rising costs, opposition, and scrutiny

The rapid construction of data centers is becoming a major political and infrastructure issue. Tech companies are expanding facilities filled with energy-intensive AI servers, while communities and lawmakers are questioning their effects on electricity grids, utility bills, water supplies, air quality, and local environments. Data centers became a bipartisan issue ahead of the 2026 midterms, and voters largely rejected further expansion in elections described in the coverage.

The Decoder

Ripple-65B Preview posts strong coding benchmark results

Ricursive Intelligence has released Ripple-65B Preview, an open-weight language model with 65 billion parameters. The AI chip design company says it was pretrained on 12 trillion tokens and reached 80.52 on TerminalBench 2.1, which it describes as the highest score from an open U.S. model. The stated total cost for research, development and training was $32 million.

TechCrunch

More than 800 Artex AI hacking tool deployments exposed online

Security researchers have identified more than 800 distinct deployments of Artex, an open-source autonomous penetration-testing tool, that were reachable online without a password. The Chinese-made software automatically searches computer systems for weaknesses and has been linked to several cyberattacks since its public release in July. It is freely available to run on private servers and has been used by organizations that cannot access advanced cybersecurity AI models from Anthropic or OpenAI.

TechCrunch

Meta launches Spaces for organizing projects with Muse agents

Meta has introduced Spaces, a new way to communicate and work with its Muse artificial intelligence agents. A Space is a separate conversation centered on a specific topic or objective. It appears in the interface sidebar, and opening it returns users to the related work.

The Verge

Meta Muse introduces customizable interactive spaces

Meta is adding a feature called “spaces” to its Muse AI assistant, allowing users to create customized, app-like experiences inside the service. A space can be designed for different purposes, including a shopping hub, a school dashboard, a music tracker or a trivia game. Users can interact with these creations directly in Muse.

TechCrunch

Midterm results could bring tighter rules for U.S. data centers

Democrats’ victories in the U.S. midterm elections may affect the technology industry, particularly the development of data centers. Voters considered 41 ballot measures related to data centers, and majorities supported additional restrictions in all but one case. Data centers also featured prominently in campaigns where no direct measure was proposed.

Ars Technica

Meta Pauses Planned AI Training Terms for Facebook Business Accounts

Meta has indefinitely postponed a planned update to its commercial terms after weeks of criticism. The proposed changes would have required users of Meta Business accounts to permit extensive artificial-intelligence training on external websites linked through Facebook. The company said its current terms will remain in effect while it continues discussions with stakeholders and works on the update.

MarkTechPost

Zyphra releases open-source FP8 attention kernels for AMD MI355X

Zyphra has released ZYFA, an open-source library of FP8 and BF16 prefill attention kernels for AMD Instinct MI355X GPUs. The Apache 2.0 project targets causal forward attention for long-context workloads, with support for FP8 e4m3, BF16, head dimension 128 and MLA shapes of 192/128. It runs on MI355X hardware with ROCm 7.2.3 and PyTorch 2.12, and provides a drop-in flash_attn_func interface.