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

Page 39 of 255

The Decoder

Google will tie school Search access to AI settings from 2027

Starting in early 2027, Google plans to keep Gemini, Gemini Notebook and Google Search turned off by default for student accounts. Schools will be able to activate each service separately, but they will not have the option of offering Google Search without its AI features.

MarkTechPost

Tencent releases WeVisOne-VL for unified visual perception tasks

Tencent’s WeChat Vision team has released WeVisOne-VL, an Apache 2.0 vision-language model available in 2B and 4B versions. The models combine detection, referring, counting, pointing, OCR, segmentation and alpha matting. A 9.6M-parameter pixel decoder produces masks and mattes without requiring a separate SAM model. When a requested object is not present, the system can return None instead of generating a box.

The Verge

Nexus Mods will require creators to disclose AI use

Nexus Mods, a repository for PC gaming modifications, is introducing mandatory disclosure of artificial intelligence use for newly uploaded mods. Mod authors will have to indicate whether AI was involved in creating their work. The platform announced the change in a broader post addressing AI-generated content on its site.

The Verge

The timing of an AI bubble burst remains unknowable

Questions about when the AI bubble will burst have become common, including among the publication’s own staff. The article’s writer says the question is asked with both excitement and concern, but offers no forecast. Instead, the writer remarks that anyone who truly knew the answer could place a life-changing, highly leveraged bet on the market, then retire—perhaps even by running a cat sanctuary.

TechCrunch

Anthropic develops FlashProve to verify AI model identity

Anthropic is developing an open-source system called FlashProve that could verify which AI model generated a particular response. The project uses what its researchers call “provable inference”: a cryptographic proof intended to provide a mathematical guarantee that an output came from a specified model.

TechCrunch

AI data centers turn to converted jet engines for faster power

AI data center developers facing years-long waits for grid connections and new turbines are installing electricity-generating units rebuilt from retired aircraft engines, the New York Times reported. The engines are modified to run on natural gas and can be deployed faster than some conventional power options, although they are not viewed as permanent power plants.

MarkTechPost

Anthropic releases FlashProve for verifiable model inference

Anthropic’s Frontier Red Team has released FlashProve, an open-source research prototype using zkSNARKs to prove that a response was generated by a specific committed model. The proof does not disclose the model’s weights. The system is designed to address a verification gap in API services, where users currently cannot independently confirm which model produced an answer or whether token output was altered.

TechCrunch

Ballistic Ventures seeks $490 million for two cybersecurity funds

Ballistic Ventures, a cybersecurity-focused venture capital firm co-founded by former Kleiner Perkins partner Ted Schlein, is seeking to raise $490 million through two new funds, according to regulatory filings. The San Francisco-based firm is targeting $360 million for its third flagship fund and $130 million for a separate vehicle called Ballistic Ventures Fund B. Neither fund had reported sales when the filings were submitted.

TechCrunch

Palantir’s strong quarter masks signs of slowing enterprise sales

Palantir reported third-quarter revenue of $2.31 billion, up 96% from a year earlier and 20% from the previous quarter. It was the company’s first quarter above $2 billion, and the result exceeded analysts’ $2.18 billion expectation. Adjusted earnings reached $0.44 per share, compared with an expected $0.41, while Palantir also raised its full-year revenue guidance.

The Verge

Hairy photos of Jeff Bezos and Jason Statham are fabricated

Images showing Amazon founder Jeff Bezos and actor Jason Statham with considerably more hair are circulating on X, Threads, and Bluesky. Both men are widely recognized for their bald or closely shaved appearance, and the source reports that this has not changed in recent days. In Bezos’s case, the altered image makes the hair look like a recent transplant.

Ars Technica

Ukraine strikes prompt Russia to consider underground data centers

Russian technology companies are seeking ways to protect and conceal data centers after Ukrainian drone strikes damaged facilities containing AI supercomputers and interrupted cloud services used by millions of Russian customers. The attacks affected infrastructure operated by Yandex, Sberbank, VKontakte and the Russian Ministry of Finance, according to the report.

MarkTechPost

Three optimizations raise Ling-3.0-flash throughput on DGX Spark

NVIDIA tested three combined techniques for mixture-of-experts inference on a single DGX Spark: Humming MXFP4 MoE kernels, an FP8 version of the language-model head, and built-in NEXTN speculative decoding. The test used Ant Group inclusionAI’s Ling-3.0-flash-fp4, a 124B-parameter model with 5.1B active parameters per token, running at batch size one on a system with 128 GB of unified memory.