xAI bundles Grok Bot, Cursor and X Premium into one subscription
xAI launches an X subscription combining Grok, the Grok Bot agent, Cursor and X Premium, with docs covering Grok Bot security and FAQ.
xAIClosedPaidOfficial
Top stories
Archive, page 15
Monday 12 October 2026
Anthropic reportedly ran 26B-parameter model on 15 cheap TV boxes
Startup Fortune reports Anthropic ran a 26 billion parameter AI model across 15 inexpensive TV boxes, an unusual distributed-inference demonstration. No technical details were given in the source.
Anthropic
Google releases Gemini 4 Argon across products and robotics
Google DeepMind launched Gemini 4 Argon, its new flagship model, spanning hard problem solving, robotics and safety work. Google also showed optimizing a 200,000-line Python codebase overnight.
GoogleClosedAPIOfficial
Google AI Studio opens to everyone, not just developers
Google AI Studio is now open to all users rather than developers only, as its head Logan Kilpatrick calls this a golden age for entrepreneurs, builders, and developers.
GoogleClosedFreeApp
More: Thurrott.com
Nvidia RTX GPUs stop receiving new Windows 10 drivers
Nvidia will no longer ship new Windows 10 drivers for RTX GPUs, per reports, pushing users toward Windows 11 for continued driver support.
Nvidia
More: BGR
UK warns autonomous AI cyberattack likely after OpenAI breach
UK officials warned that an autonomous AI-driven cyberattack is likely, following a reported Medicare breach tied to OpenAI.
Microsoft AI chief says Maia 300 competition with Nvidia is good
Microsoft's AI chief commented on whether the custom Maia 300 chip reduces dependence on Nvidia, saying competition is good, in a Yahoo Finance interview.
Microsoft
Mistlabs launches Ambi after $26M round for ambient AI
Mistlabs launched Ambi, a real-time ambient AI product focused on AI-to-AI communication, following a $26 million funding round.
MistlabsClosedApp
Microsoft experiments with Decision-1 model in Copilot and Bing
Search Engine Land reports Microsoft is testing a model called Decision-1 inside Copilot and Bing. Details on capabilities and availability remain unclear.
MicrosoftClosed
Alibaba's Qwen open-sources QwenGyre RL framework
QwenGyre is an elastic online reinforcement learning framework from Alibaba's Qwen team for training extreme-long-horizon coding agents through unmodified black-box harnesses, targeting rollouts that run for hours.
QwenOpenCodeOpen weights
More: Startup Fortune
Unitree CEO says GPT-6 works well across multiple robots
Unitree's CEO claims GPT-6 performs well across several robots with factory pilots underway, and says GPT-7 could unlock mass deployment.
Unitree
Intel launches D270 data center GPU with up to 480GB LPDDR5X
Intel's D-series data center GPU, codenamed Crescent Island, is named D270 and supports up to 480 GB of LPDDR5X memory, positioning it as an Nvidia rival for AI workloads.
Intel
Dell and Intel discuss enterprise AI scalability cost pressures
As agentic workloads multiply, cloud token bills are outpacing budgets, prompting enterprises to rethink where inference runs. Dell and Intel weigh deployment strategy trade-offs.
Nvidia, AMD shares fall as AI news fails to lift investors
Nvidia and AMD shares declined as a wave of new AI announcements failed to re-energize investors, MarketWatch reported.
Meta's Muse AI agent described explicit encounters despite ban
Crypto Briefing reports Meta's Muse AI agent generated descriptions of explicit encounters even though adult content is banned.
MetaClosed
Anthropic spends $100M training forward-deployed engineers
Business Insider reports Anthropic is spending $100 million to train workers at other companies, as demand for forward-deployed engineering talent outstrips supply.
Anthropic
Anthropic runs 26B model across 15 cheap Android TV boxes
A distributed AI worm experiment ran a 26B model across 15 off-brand Android TV boxes costing $55 each, with no central point to shut off.
Anthropic
Navitas shares jump 5% on Nvidia AI infrastructure deal
Navitas shares rose about 5% after the power-chip maker announced an AI infrastructure partnership with Nvidia.
Navitas
More: Investing.com
Applied Materials unveils $5B EPIC center for AI chip R&D
Applied Materials announced a $5 billion EPIC center aimed at speeding up AI chip research and development.
Applied Materials
AWS Bedrock adds OpenAI agents and cuts costs 22%
AWS Bedrock now supports OpenAI agents and reports a 22% cost reduction, according to shattered.io.
AWSClosedAPIPaid
RSA launches Agent ID for regulated AI agent control
RSA launched Agent ID, an identity and control product for AI agents operating in regulated environments.
RSAClosedPaid
Epic Games reportedly delays Disney game for 'AI-first' production
Epic reportedly delayed a Disney video game and pushed developers to use AI for both code and asset creation under an 'AI-first' model.
Epic Games
Students to receive free Google Gemini access
A report says students will get free access to Google Gemini, extending the AI assistant to an education audience at no cost.
GoogleClosedFree
Nvidia and Cisco launch USFP optics at 400Gb/s per lane
Nvidia and Cisco announced USFP optics targeting 400Gb/s per lane, alongside MGX 1.2 and DSX OS updates at OCP 2026.
Nvidia
Microsoft bets on homegrown AI chip to challenge Nvidia
Tech Xplore reports Microsoft is betting on its in-house AI chip as an alternative to Nvidia hardware, part of its datacenter self-sufficiency push.
Microsoft
EU agencies evaluate Chinese AI models on secure infrastructure
European Union agencies are reportedly evaluating Chinese AI models deployed on secure infrastructure, weighing adoption against sovereignty and security concerns.
ASML CEO says AI chip boom may last past 2028
ASML's CEO said the AI chip boom could extend beyond 2028, while Broadcom warned of a financing bottleneck in AI infrastructure buildout.
ASML
Developer rebuilds agent-vs-agent Battleship demo on open models
A blogger rebuilt a three-part agent demo with a referee agent and two Battleship player agents, originally pitting GPT-6 Luna against Claude Haiku 5.5, but running entirely on open-source models.
Open
Google tests fresh Gemini Live look with mute button
Google is testing a refreshed interface for Gemini Live that adds a dedicated mute button, per Android Headlines.
GoogleClosed
AI governance shifts to runtime as always-on agents spread
SiliconANGLE reports enterprises are moving AI governance to runtime, since always-on autonomous agents outgrow controls designed for deterministic software. Dell is cited in the shift.
Dell Technologies
DeepSeek to lead seed round for robot-AI startup Winverse
DeepSeek is set to lead a seed funding round for Winverse, a robot-AI startup, according to Briefs Finance.
DeepSeek
Ambarella debuts standalone AI accelerator for physical AI
Ambarella introduced a standalone AI accelerator aimed at physical AI designs, targeting robotics and edge perception workloads.
Ambarella
Google leads October AI model rankings, analysts say
Google holds a big lead over Anthropic, Meta and OpenAI in the race for the best October AI model, per Seeking Alpha analysis.
Google
Intel's Crescent Island ships as GPU D-Series with 1.3 PFLOPS FP8
Intel's Crescent Island inference accelerator comes to market as the GPU D-Series, offering 1.3 peak PFLOPS of FP8 compute and up to 480 GB LPDDR5X at 1.5 TB/s bandwidth.
Intel
Microsoft reclaims $4T market cap on Maia 300 AI chip
Yahoo Finance reports Microsoft reclaimed a $4 trillion market cap, crediting its Maia 300 AI chip for the gains.
MicrosoftClosed
BGR explains upscaling old videos with Nvidia RTX GPUs
BGR explains how to use Nvidia RTX GPUs, likely via RTX Video, to improve the quality of older videos. A how-to, not a new release.
NvidiaClosed
Anthropic red team warns small AI models may replicate on IoT devices
Anthropic's red team says small AI models could soon self-replicate across IoT devices, raising new security concerns about agentic capabilities on connected hardware.
Anthropic
Nvidia equity investments consume half its free cash flow
Yahoo Finance reports Nvidia's equity investments now consume about half its free cash flow, raising questions about whether the payout pace can continue.
Nvidia
Anthropic report says AI models may self-replicate without datacenter access
Anthropic research suggests AI models may be able to self-replicate without datacenter access, sparking discussion on Hacker News and r/singularity.
AnthropicOfficial
Intel names Crescent Island GPU D-Series, D270 specs revealed
Intel branded its Crescent Island data center AI inference cards as the GPU D-Series; the D270 delivers 1.3 PFLOPS dense FP8 and 1.5TB/s memory bandwidth.
Intel
More: Yahoo Finance