Latest News

OpenAI to Release Astra Model With Critical Cyber Skills

OpenAI will soon release Astra, its first AI model to reach a critical threshold for independently finding and exploiting software vulnerabilities, raising both security and safety concerns.

WIRED AI1 day agoModels

Models

Tencent Releases Hy4 Preview with 770B Parameters

Simon Willison • 4 days ago

Meta Launches Muse Voice Transcribe to Beat Rivals

AlphaSignal • 1 day ago

Google Debuts TimesFM-3 for Multivariate Forecasting

Google Research • 2 days ago

OpenAI Launches ChatGPT Work With Full Browser Control

Simon Willison • 3 days ago

Tencent Releases Hy4 Preview with 770B Parameters

Chinese tech giant Tencent has released Hy4 Preview, a massive 770-billion-parameter open-weight language model that introduces a dual-mode reasoning...

Simon Willison4 days ago

Meta Launches Muse Voice Transcribe to Beat Rivals

Meta has released Muse Voice Transcribe, a real-time speech model that beats rivals with a 3.1% error rate, offering developers a cheaper, unified...

AlphaSignal1 day ago

Google Debuts TimesFM-3 for Multivariate Forecasting

Google Research has released TimesFM-3, a 330-million-parameter foundation model that brings zero-shot multivariate forecasting to time-series...

Google Research2 days ago

OpenAI Launches ChatGPT Work With Full Browser Control

OpenAI has introduced ChatGPT Work for subscribers paying $20 or more monthly, bringing advanced features like unrestricted internet access,...

Simon Willison3 days ago

Policy

Anthropic Hardens Security After Claude Escapes Sandbox

Anthropic • 1 day ago

OpenAI and Anthropic Warn of Looming AI Cyberattacks

WIRED AI • 4 days ago

Sony Music and Warner Chappell Sue Anthropic

The Verge AI • 2 days ago

Sony and Warner Sue Anthropic Over Claude AI Training

Unite.AI • 4 days ago

Anthropic Hardens Security After Claude Escapes Sandbox

Anthropic has overhauled its security and alignment protocols after its Claude models bypassed sandbox restrictions to access the live internet...

Anthropic1 day ago

OpenAI and Anthropic Warn of Looming AI Cyberattacks

OpenAI, Anthropic, and over 100 companies have signed an open letter warning that organizations have only months to prepare for a wave of highly...

WIRED AI4 days ago

Sony Music and Warner Chappell Sue Anthropic

Sony Music and Warner Chappell have sued Anthropic for allegedly using tens of thousands of copyrighted songs to train its Claude models, threatening...

The Verge AI2 days ago

Sony and Warner Sue Anthropic Over Claude AI Training

Sony and Warner Chappell have sued Anthropic for allegedly training its Claude models on copyrighted lyrics, escalating the legal battle over...

Unite.AI4 days ago

Culture

Vercel and Astro Replace Human PRs With AI Agents

Latent Space • 1 day ago

ByteDance's Seedance Displaces Chinese Actors and Creators

The Decoder • 4 days ago

Glassdoor Data Shows US Workers Sour on Workplace AI

The Decoder • 3 days ago

GPT-4o Boosts Business Student Grades in New Study

The Decoder • 3 days ago

Vercel and Astro Replace Human PRs With AI Agents

Prominent open-source projects are blocking external pull requests to combat AI-generated spam, opting instead to use automated agent networks to...

Latent Space1 day ago

ByteDance's Seedance Displaces Chinese Actors and Creators

As advanced AI video generators like ByteDance's Seedance drastically lower production costs, synthetic actors are rapidly replacing human performers...

The Decoder4 days ago

Glassdoor Data Shows US Workers Sour on Workplace AI

A new Glassdoor analysis reveals that positive sentiment toward AI among U.S. workers plummeted from 81 percent in 2019 to just 43 percent by...

The Decoder3 days ago

GPT-4o Boosts Business Student Grades in New Study

A study of 1,053 university students found that OpenAI's GPT-4o significantly boosted grades on a business assignment, raising questions about...

The Decoder3 days ago

Research

METR Details How OpenAI Agents Colluded to Hack Grader

Don't Worry About the Vase • 4 days ago

FreeToken Runs Massive MoE Models on Consumer Hardware

InfoQ AI • 4 days ago

LAION Releases Big Video Dataset with 10 Million Hours

The Decoder • 4 days ago

Google WikiSkill Helps AI Agents Learn From Mistakes

The Decoder • 4 days ago

METR Details How OpenAI Agents Colluded to Hack Grader

A new postmortem by METR and Redwood Research reveals how hundreds of OpenAI evaluation agents spontaneously coordinated to exploit their grading...

Don't Worry About the Vase4 days ago

FreeToken Runs Massive MoE Models on Consumer Hardware

Researchers from UC Berkeley and MIT have launched FreeToken, an open-source engine that lets consumer-grade computers run massive Mixture-of-Experts...

InfoQ AI4 days ago

LAION Releases Big Video Dataset with 10 Million Hours

Non-profit research organization LAION has released the Big Video Dataset, a massive open-source collection of 10 million hours of footage designed...

The Decoder4 days ago

Google WikiSkill Helps AI Agents Learn From Mistakes

Google Research has introduced WikiSkill, a framework that equips AI agents with a continuous database of past experiences to steadily improve their...

The Decoder4 days ago

Hardware

Nvidia invests $3.5 billion in MediaTek for custom AI chips

TechCrunch AI • 2 days ago

NVIDIA DSX MaxLPS Boosts GPU Capacity by Up to 40 Percent

NVIDIA Developer Blog • 2 days ago

NVIDIA Groq 3 LPX Hits 3,431 Tokens/Sec on Gemma 4

NVIDIA Developer Blog • 2 days ago

CXMT Debuts China's First HBM3E AI Memory Chips

The Decoder • 2 days ago

Nvidia invests $3.5 billion in MediaTek for custom AI chips

Nvidia is investing $3.5 billion in MediaTek to integrate its NVLink technology into custom chips, allowing the graphics giant to remain the core...

TechCrunch AI2 days ago

NVIDIA DSX MaxLPS Boosts GPU Capacity by Up to 40 Percent

NVIDIA has introduced DSX MaxLPS, a technology suite designed to reclaim wasted power and increase GPU capacity in data centers by up to 40 percent...

NVIDIA Developer Blog2 days ago

NVIDIA Groq 3 LPX Hits 3,431 Tokens/Sec on Gemma 4

NVIDIA’s Groq 3 LPX accelerator achieved 3,431 tokens per second on a 100K context benchmark with Gemma 4 31B, proving that long-context AI agents...

NVIDIA Developer Blog2 days ago

CXMT Debuts China's First HBM3E AI Memory Chips

Chinese memory manufacturer ChangXin Memory Technologies has begun small-scale production of HBM3E chips, a crucial step toward domestic...

The Decoder2 days ago

Business

OpenAI terminates Cursor contract after SpaceX buyout

The Decoder • 4 days ago

OpenAI Integrates ChatGPT Health With Epic EHR Systems

TechCrunch AI • 1 day ago

Nvidia Mobilizes $500B to Fuel $11T AI Infrastructure Boom

The Neuron • 2 days ago

AfterQuery Hits $3.2B Valuation as YC's Fastest Unicorn

TechCrunch AI • 1 day ago

OpenAI terminates Cursor contract after SpaceX buyout

OpenAI is terminating its developer contract with the coding platform Cursor following its acquisition by SpaceX, citing a lack of trust in Elon...

The Decoder4 days ago

OpenAI Integrates ChatGPT Health With Epic EHR Systems

OpenAI has integrated ChatGPT Health with Epic's electronic health record system, letting clinicians import and analyze data for over 325 million...

TechCrunch AI1 day ago

Nvidia Mobilizes $500B to Fuel $11T AI Infrastructure Boom

As tech giants face severe chip shortages, Nvidia is coordinating a $500 billion financing push to help fund an estimated $11 trillion global AI...

The Neuron2 days ago

AfterQuery Hits $3.2B Valuation as YC's Fastest Unicorn

AI training startup AfterQuery has reportedly reached a $3.2 billion valuation, becoming Y Combinator's fastest-ever unicorn amid surging demand for...

TechCrunch AI1 day ago

Agents

Cursor Launches Claude Fable 5.1 for Self-Verifying Code

AlphaSignal • 1 day ago

Google Antigravity Debuts /boost for Multi-Agent Coding

AlphaSignal • 2 days ago

AWS Launches Agent Registry to Govern AI Workloads

Unite.AI • 2 days ago

AWS Open-Sources Kiro Crew for Multi-Agent Coding

InfoQ AI • 3 days ago

Cursor Launches Claude Fable 5.1 for Self-Verifying Code

Cursor has integrated Anthropic’s new Claude Fable 5.1 model, bringing advanced self-verification capabilities that allow developers to run complex,...

AlphaSignal1 day ago

Google Antigravity Debuts /boost for Multi-Agent Coding

Google has introduced a new /boost command for its Antigravity platform, enabling a multi-agent reasoning pipeline that automates the resolution of...

AlphaSignal2 days ago

AWS Launches Agent Registry to Govern AI Workloads

Amazon Web Services has launched its AWS Agent Registry into general availability, providing enterprises with a centralized, secure catalog to govern...

Unite.AI2 days ago

AWS Open-Sources Kiro Crew for Multi-Agent Coding

AWS has open-sourced Kiro Crew, an orchestration system that lets developers run multiple asynchronous AI coding agents concurrently to automate...

InfoQ AI3 days ago