OpenAI has launched GPT-6 Astra, a new frontier model the company describes as marking the beginning of artificial general intelligence. Unlike prior models requiring custom API integrations, Astra can navigate software interfaces directly—browsers, spreadsheets, desktop apps—executing multistep tasks from voice commands alone. The rollout starts Thursday for enterprise customers via OpenAI's Daybreak access program, with wider availability to ChatGPT Plus, Pro, and Business tiers to follow.
According to The Information, OpenAI's upcoming Astra model employs a technique called recurrent depth or opaque recurrence, letting it think outside the step-by-step chain-of-thought format used by most reasoning models. Reports indicate the technique's use in Astra is currently limited, but its introduction has drawn sharp criticism from AI safety researchers who worry about losing insight into model reasoning.
A developer documented the process of bringing up the Linux kernel on a new platform by writing a lightweight emulator for a RISC-V CPU rather than using real hardware. The emulator, built in roughly 2000 lines of C++ and implementing the RV32IMA instruction set, provides just enough hardware simulation—MMU support, RAM, an SBI interface, a system timer, and a WD8250-based UART—for Linux to boot. The full source is published on GitHub under WerWolv/riscv-emulator.
Joe Drygas, Vice President of AT&T's Global Enterprise Business, describes how he rose through the company over 25 years by treating each new role as requiring a reinvented version of himself, comparing the approach to how elite athletes adapt their training as their roles evolve. He also credits deliberate scheduling of personal routines, like workouts and sleep, with reducing uncertainty and improving his leadership clarity.
Billionaire investor Leon Black skipped a scheduled sworn deposition before the House Oversight Committee on Thursday, instead filing a federal lawsuit seeking to block subpoenas tied to the panel's investigation into Jeffrey Epstein. Committee leaders responded by saying they would consider holding Black in contempt of Congress.
Crystal Temptations issued a recall of its Chocolatey Eyeballs product after the FDA flagged a safety concern, reportedly tied to undisclosed allergens that could pose serious health risks. The candy was sold at major retailers including TJ Maxx and Marshalls across 13 states. The FDA has advised consumers who purchased the product to destroy it rather than consume or return it.
MikroTik has released the CRS804-4DDQ-hRM, a compact desktop switch offering four QSFP56-DD ports for a combined 1.6Tbps of 400GbE bandwidth. It carries a $1,295 list price but typically sells for around $1,100, and the reviewer has been running it in RDMA backends for NVIDIA GB10 and GB300 clusters for months.
Meta released Muse Spark 1.3, an upgrade to its AI coding and agent model, touting benchmark scores that mostly come from a 'max' reasoning configuration still undergoing safety testing and unavailable through any API provider. The version actually shipping to developers via Muse Code and the Meta Model API uses the older 'xhigh' reasoning setting, which scores meaningfully lower on tests like GDPval-AA v2, OSWorld 2.0, and JobBench. Meta disclosed both sets of results in its evaluation report, but its marketing emphasized the higher, currently inaccessible numbers.
Nvidia has agreed to acquire AI platform Hugging Face for roughly $12.9bn, one of its largest acquisitions to date. Hugging Face, used by over 18 million developers and hosting more than three million AI models, operates as a major hub for sharing open-source AI tools. Nvidia said the platform will remain open and won't require users to adopt its chips or services.
At IFA 2026, Singapore's Lingverse introduced iKairos, a device that functions as a desktop AI companion but can be removed from its robot base and worn as a neck pendant. The device gathers contextual data from both settings and feeds it into Lingverse's LifeOS to generate personalized suggestions, such as recommending activities based on recurring conversation topics.
Nvidia has officially detailed its RTX Spark chip, branding it the N1X and confirming it will ship in laptops and mini PCs starting in October. The lineup includes a top config pairing a 20-core Grace CPU with a 6,144-core Blackwell GPU and up to 128GB of unified memory, plus a lower-tier laptop-only version with 18 CPU cores, 5,120 CUDA cores, and up to 32GB of memory. Lenovo and Acer have already unveiled devices using the chip ahead of IFA Berlin, with more OEMs expected to follow.
Lenovo has published detailed specifications for two Nvidia RTX Spark-based laptops: the Yoga Pro 9n clamshell and the Yoga 9n 2-in-1 convertible. Both use Grace CPUs with up to 20 cores and Blackwell RTX graphics with up to 6,144 cores, but differ in memory, storage type and screen size, with the Pro model offering higher-end specs overall.
GoKawiil