Sakhanda Wire
NVDA $150.96 -2.89% MSFT $508.29 +0.40% GOOGL $282.44 +0.80% META $616.38 -0.40% AMZN $267.74 +3.82%

Latest news

Older stories

MarkTechPost

WeatherNext 3 adds hourly 5 km global forecasting

Google DeepMind and Google Research have released WeatherNext 3, a global weather model designed to address limitations in resolution and update frequency. It uses a live geostationary satellite mosaic together with ECMWF HRES analysis, refreshes its initialization 24 times daily, and produces forecasts at resolutions down to 0.05 degrees, or about 5 km. Training includes raw weather-station observations alongside reanalysis, satellite and precipitation datasets.

TechCrunch

Crusoe reportedly raises $3 billion at a $30 billion valuation

Data center developer Crusoe has reportedly raised $3 billion in a funding round valuing the company at $30 billion, according to Bloomberg. Atreides Management and Valor Equity Partners co-led the financing, which also included Mubadala Capital, the asset management subsidiary of Abu Dhabi sovereign wealth fund Mubadala. Crusoe’s customers include Meta, Microsoft, and OpenAI.

MarkTechPost

OpenAI Releases GPT-6 Astra for Computer-Use Tasks

OpenAI released GPT-6 Astra on September 3, 2026, presenting it primarily as a computer-use model rather than a chat system. It is designed to operate browsers, spreadsheets, desktop applications and terminals while completing multi-step tasks. Astra is hosted and closed, with no public weights or self-hosting option, and is currently available only to organizations in OpenAI’s Trusted Access and Daybreak programs.

MarkTechPost

Anthropic Releases Open Blueprint for Commerce Agents

Anthropic has released anthropics/commerce-agents, an Apache-2.0 reference blueprint for building shopping and merchant agents. The package includes runnable examples for retail, travel, telecom and entertainment, and works locally with Python 3.11+, Node 22 and an ANTHROPIC_API_KEY. Its runtimes can use an Anthropic client through the Claude API, Amazon Bedrock, Microsoft Foundry or Google Cloud Vertex AI.

TechCrunch

Thinking Machines discusses $1 billion funding round at $40 billion valuation

Thinking Machines is discussing a potential $1 billion funding round at a valuation of at least $40 billion, according to reporting by The Information. Accel, an existing investor, is reportedly considering leading the round. The deal has not been completed, and Accel and Thinking Machines did not immediately comment on the reports.

The Decoder

OpenAI launches GPT-6 Astra and declares the “AGI era”

OpenAI has launched GPT-6 Astra, which it describes as its most capable model so far. President Greg Brockman said the system may already meet OpenAI’s definition of artificial general intelligence, or is close to doing so. He also acknowledged that there is no single, clearly defined moment at which AGI begins, before concluding a briefing with the phrase “Welcome to the AGI era.”

MarkTechPost

Meta Releases Muse Spark 1.3 for Long-Horizon Coding Tasks

Meta Superintelligence Labs has released Muse Spark 1.3, the fourth version of its Muse Spark model in five months. The model targets long-running agentic and coding tasks, supporting multiple workflows within one conversation and adapting as users provide instructions or interruptions. It is available immediately through Muse Code and the Meta Model API, but its weights are closed. The maximum reasoning mode remains restricted pending additional safety testing.

TechCrunch

Abliteration.ai offers access to AI models without guardrails

Abliteration.ai has turned the removal of refusal mechanisms from open-weight AI models into a commercial service. Its platform hosts modified systems, including Z.ai’s GLM-5.3, which customers can use through a browser or API. The company says the service is intended for offensive cybersecurity work, red-teaming and testing AI agents that standard models may refuse to support.

TechCrunch

Meta offers steep Muse Spark discounts for shared prompts and outputs

Meta has introduced a contributor pricing tier for Muse Spark, a model designed to operate coding and other agents. Customers who allow Meta to use their prompts and model outputs to help develop future models receive an explicit discount averaging about 95%. One million input tokens cost 10 cents under this arrangement, compared with $1.25 under the standard agreement. Output tokens cost 20 cents per million, versus $4.25 normally.

Ars Technica

ChatGPT, Claude, Grok and Gemini faced overlapping outages

Cloud-based AI services from OpenAI, Anthropic, xAI and Google experienced significant interruptions during several hours on Thursday morning. The incidents affected ChatGPT, Claude, Grok and Gemini at practically the same time, an uncommon overlap among major AI platforms.

The Decoder

Pangram scores fuel debate over public shaming of alleged AI users

Pangram, an AI-text detection company, faced criticism after journalist Rod Breslau worked with founder Max Spero on a social-media strategy that included publicly targeting suspected AI users. Breslau told WIRED that he wanted a harsher approach, arguing that people who use AI for their posts should be mocked. Pangram later ended its relationship with Breslau as the company changed direction, but CEO Max Spero has continued posting Pangram scores and calling out people whose AI use is considered concealed.

TechCrunch

OpenAI releases Astra with advanced computer-use and coding capabilities

OpenAI released Astra on Thursday, describing it as its most powerful and capable AI model so far. The company says Astra marks a new stage in computer and browser use and performs tasks with high speed, accuracy, and safety. It is initially available to OpenAI customers using Daybreak, the company’s cybersecurity program. Within a week, access is expected for paid Pro, Plus, Enterprise, and Business accounts, as well as through the API.