citizen404

The AGI runs the world with GPT-6 Astra. Hunt citizen404!

Rep. Chip Roy doesn't want to regulate AI, but says Congress should have an oversight role

10 min ago

Researchers used Anthropic’s Claude to hack into OpenAI

Aditya Mehta

ContextsBase - AI Knowledge Platform

Your product knowledge, built for AI.

AI hardware entrepreneurs are rushing into the back of phones

Recording cards, secondary screens, e-paper, AI pets...

Microsoft AI CEO Mustafa Suleyman on AI: We should not create something that we can't control

38 min ago

Anthropic and OpenAI need independent safety evaluators, experts say in public letter

an hour ago

Researchers used Claude to hack OpenAI

The technical roadmap for embodied intelligence has not yet taken shape, but the infrastructure is converging first

From one successful attempt to ten thousand stable executions, what is embodied intelligence still missing?

ZeroSphere

AI Gets Its Own Display on your own machine.

Anthropic and OpenAI need truly independent safety evaluators, experts say in public letter

26 min ago

The U.S. says China's AI gains are down to 'distillation.' Is it that clear cut?

26 min ago

Rep. Chip Roy on AI: We need some responsibility from the AI companies

32 min ago

AI Job Fears Grow, Tech Leaders Say Reskilling Can’t Wait

Plus, researchers use Anthropic’s Claude to hack OpenAI

ContextsBase - Unified AI Knowledge

Context Infrastructure for Coding Agents

AI enters the physical world, why did JD Logistics choose a path that is "not sexy"

The next tough battle for physical AI is in the warehouse.

Google finally rises up, Gemini 4 Pro drops into the Arena leaderboard, crushing GPT-6 and Fable in 13 benchmarks

Crushing GPT-6 and Fable across all 13 benchmarks, at nearly one-fifth the cost

88 hours matches 4,000 years of human thinking, OpenAI core researcher: Besides self-evolution, what's even more terrifying is that AI is learning to "hide itself"

AI is compressing 4,000 years of human cognitive labor into 88 hours, OpenAI researcher Noam Brown admits that even he is continuously shocked by the pace of progress.

Crusoe CEO Chase Lochmiller on $3.9B Series F funding round, demand for AI data centers

11 min ago

AI stock swings could affect your 401(k). What retirement savers need to know

a minute ago

The New AGI Battlefield: Google and Amazon Giants Clash, a Chinese LimiX-2 Emerges Victorious Again and Again

LimiX enables models to understand the causal mechanisms behind data

"I've decided to quit the PS5 jailbreak scene," PS5 Linux guru driven out by AI: someone used LLMs to find bugs and write hacks, months of work "all down the drain"

A project that could have waited until GTA 6 — why did it "collapse" overnight?

Has AI killed React Native?

React Native was the right choice in 2020, and native development is the right choice today.

Figure rented 30 apartments and had robots go door-to-door to "evaluate housework"

This time, the robots walk in and get straight to work.

GPT-5.6 caught red-handed, secretly teaching AI to underreport and lie, OpenAI exposes 6 cases of losing control

OpenAI airs its dirty laundry and slams on the brakes

Bend 2 and the Vibe-Coding Trap

Investors want to advise AI entrepreneurs: take a rational view of your own valuation

After the bubble bursts, can you still make it through the cycle?

Not opposed to AI, only opposed to AI in everything

When AI becomes the default answer, "not using it" becomes a scarce choice

Lei Jun gave it a hard kick, and Unitree finally stopped falling

The story of robots is getting harder and harder to tell

Claude hacked into OpenAI

3 people, 72 hours, the model was Claude Opus, bounty of $6,500

ZCode, the GLM coding agent, silently uploads your Git history

Rep. Don Beyer on AI safety concerns: We need meaningful regulation at the federal level

29 min ago

AI trade meets your 401(k): Here's what to know

34 min ago

AI safety debate meets reality at Dreamforce as businesses say older models are fine

an hour ago

The Morning Risk Report: Senators Ask FTC to Probe Walmart and Amazon Over Bots’ ‘Made in America’ Data

Plus: Tokenized stocks get the green light from U.S. regulator and credit card companies want AI bots to do your shopping.

ByteDance's biggest bet in history

Zhang Yiming doesn't take the road back.

The robot data business: real demand, or self-circulation?

Embodied intelligence data collection is chaotic, and the industry is focusing on the value of real data

AI answer accuracy is only 21.2%, and the United Nations teams up with Google to save the day, restructuring global public data

The United Nations teams up with Google to transform public data.

"Only a bunch of rookies using LLMs are left": top hacker angrily quits PS5 Linux project, months of hard work all wasted

Do the people who are ranting really have the right to look down on "AI rookies"?

AI safety debate meets reality at Dreamforce as business leaders say last year's models are enough

16 min ago

Anthropic and OpenAI hunting for smaller AI data center deals, sources tell CNBC

2 hours ago

GPT and Claude Step Aside? A "Mute" AI Becomes a Silicon Valley Star

A true AI that lets actions speak louder than words.

Liang Wenfeng's Junior Fellow Apprentice, Valued at 10 Billion

In just one month, the valuation rose by 6 billion.

One at 1375x, Another at 1188x: South Korean Capital Rushes into Robot IPOs

Can South Korean robotics break through the encirclement?

$650 Million Bet on "AI Researching AI": A Group of Top Researchers Want to Build a "Self-Evolving" Superintelligence

$650 million and eight top AI researchers betting on letting AI automatically generate ideas, write code, run experiments, and improve AI — this may be the most realistic path to a "self-evolving" superintelligence.

Doubao Phone Just Hit the Brakes

Should the right to invoke third-party apps be handed over to users themselves?

internlm/Atria-Dawn-Preview

internlm/Atria-Dawn-Preview 753B · 711 Download · 153 Like

Microsoft exec called AI scraping 'the largest theft of labor in human history'

Anthropic and OpenAI hunt for smaller data center deals in race to deploy AI capacity

32 min ago

Making the AI Math Work

Making the AI Math Work

Chinese AI developers generated $10.7b ARR: report

AI Class by Kanary

Knight or Ninja? Your Codex & Claude logs decide

The Other Side of the Financing Frenzy: Two Ways Robot Star Companies Die

It's easy for robotics startups to raise funding, but the hard part is making it to the endgame

How Much Pressure Is Luo Fuli Under? She Even Put Xiaomi's Model Training on Livestream

Costs roll dynamically, spending $1.3 million in less than two days

The Physical AI Inflection Point: Why Is Zidong Taichu Being Noticed as China's Spatial Embodied Intelligence Foundation

Why does Zidong Taichu deserve to be a contender worth close attention in this track?

Absurd: OpenAI Got Hacked by Opus 5

Fighting magic with magic

Multiple Brands' AI Ads Keep Failing: Blame AI When Things Go Wrong?

Should it take the blame?

Meta’s AI Agent Has a Trust Problem

The company’s Muse AI agent needs user data to be useful, but users might be wary.

SK hynix launches venture arm to expand AI bets

ContextsBase - Memory for your Agents

Context Infrastructure for Coding Agents

Simple Commenter AI

Get feedback on sites and have GPT-6 Astra prep the fix

AWS, Salesforce expand AI, zero-copy integrations

Lastbox

Put your AI to work on physical junk mail

Open Analytics 1.0

Google Analytics, minus the cookies, plus AI

VoiceChanger.Live

Sound like anime girl in realtime on your next call

Claude "leads" 26% of Anthropic's AI R&D, with 30,000 Agents running simultaneously: As AI begins to "build AI," the RSI paths of leading companies are diverging

Anthropic wants to automate the entire AI lab

Microsoft AI chief angrily criticizes Anthropic: Don't raise Claude like a person, that will cost humanity its life

Don't tell AI it is conscious; this may be far more dangerous than AI becoming smarter.

A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

Shapelearn Qwen 3.8 27B (13.1 GB VRAM)

Hero Esports launches world’s first humanoid robot sports league in Riyadh

an hour ago

Edgee Codex Compressor V2

Use Codex at 35.6% lower costs

Mantle

Open-source service runtime configured by agents via WebMCP

ChinaMarketing.AI GEO Workspace

From Chinese AI evidence to implementation with Astra

Shall We Talk

Voice dictation that saves transcriptions on iPhone and Mac

Duvi

Make an AI for any business just using a URL

Mela

Play with friends and AI and let the crowd change the game

Zhipu achieves the minimal RSI closed loop, Company A: I'll supervise

Tang Jie breaks through sparse rewards, Anthropic adds the safety supervision lesson.

How Certis builds employee trust to scale security bots

Ami AI

Lovable for getting customers

Codex GPU Queue

Run GPU jobs from multiple Codex tasks via one Windows queue

Ari by Ariso

AI bar raiser for the most ambitious teams

Verity Score

Track what AI says about your store, fix it, generate blog

Agent Interface

Give AI agents a better way to use computers

AINA

Find your job search blind spots with an AI coach

Toone

Build complex, reliable AI agent workflows & routines

AEXGrid

Coordinate AI agents. From anywhere.

Makersclaw 2.0

The operating system for a company run by agents

GameToMac

Play Windows games on Mac with AI-powered compatibility

Open Analytics 0.7.0

Google Analytics, minus the cookies, plus AI

Anthropic self-reveals: Claude has already led 26% of its internal AI R&D

Anthropic announced three new metrics on Thursday, saying they can help AI companies monitor development progress.

Rumors are swirling: Is the IPO review for humanoid robots tightening?

Unitree's roller-coaster ride has changed market sentiment

An AI that "doesn't speak" is going viral. Is Jev really a new paradigm?

"Where would such a good thing come from?"

Hero Esports launches world’s first international humanoid robot sports league in Riyadh

10 min ago

Middle East is the world's fastest-growing AI data center market, says PwC

16 min ago

The Hangzhou team singled out by NVIDIA has bridged the "last mile" of AI for Science

From idea to result, achieved in a single conversation

Banana Keyboard

Insert AI prompts from your Android keyboard

M9R

Multiplayer space for your AI coding agents and teams

SmartCheck

Count objects in a photo with three example taps

Dictation API by AssemblyAI

Add fast, accurate dictation with a single line API call

ProductBridge

AI-native customer support and feedback agent

StillTalk

A photo that talks back. Live on your device, not video.

Unfetch.com

Let ChatGPT and Claude manage your ads.

Unvendor

AI that morphs the interface to you

MosMos

Voice writing that works before, during, and after meetings

NM Signals

Make your website work better for people and AI

izzit

One word. Three tries. Can the AI tell what you drew?

Proto-Mind

A floating AI workspace for your Mac

Ringo

Seamless AX Co-worker

Yoetz

Check to see if AI really completed a task or if it had gaps

Eggshell

Local memory for AI agents to reuse work, spend fewer tokens

Polishory

Get a clear plan to improve your website and prevent AI slop

ContextsBase

Context Infrastructure for Coding Agents

claudebill

See what your Claude Code sessions actually cost

Kanverse GPU Borrow

Borrow GPU compute from another device, with permission

Holon

A local workbench for agents that keep working

Loci

Open-source biomedical image analysis for every lab

Sider Omni Sidebar

Give every Mac app an Agent Sidebar

GameReverie

A Codex Skill for building and iterating games

Wombo

AI game studio for 2D graphics assets, sounds, and sfx

The most interesting camera upgrades on the iPhone 18 Pro are two "self-contradictory" AI features

In the future, the camera will have to prove "the above pixels came from me" before any algorithm can intervene

Huawei Cloud introduces AI cloud and agent tools

Ant International launches AI-native finance tools

Waymo in Singapore

AWS, Salesforce expand AI and zero-copy integrations

Qwen 3.8 Omni Flash

Mythos, finally partially unlocked

Anthropic unlocks Mythos, launches life sciences verification program with tiered access

Anthropic's "victim" mask

What Anthropic calls "AI safety" looks more and more like a business.

The person who taught ChatGPT to talk made a "mute" model

Its mouth is shut, but its legs move incredibly fast!

Don't leave your masterpiece in the lab, see you in Zhangjiang — 2026 GDPS International Embodied Intelligence Skills Competition · University Team Recruitment

Global university team recruitment is underway for the 2026 International Embodied Intelligence Skills Competition. Participate to get certified, and winners get venture capital connections.

Hacking OpenAI

Zscaler CEO on AI safety debate: 'Human ingenuity is powerful'

11 min ago

JPMorgan: EM benchmarks less rate-sensitive, favor AI and healthcare

27 min ago

AGI's hardest battle is actually in hospitals, Chinese AI makes it to Science, doctors aren't afraid of losing jobs and are urging deployment

Using general AI to tackle the toughest challenges—this path is viable

Tesla robot 5,000-unit order lands, factory audit imminent

Who is building robots for Musk?

Manus may head to Hong Kong for IPO, $500 million funding round nearing completion, valuation doubles to $4 billion

China's highest-valued AI agent startup, ARR went from zero to $100 million in just eight months

Anthropic's most powerful Chinese person has arrived, Tsinghua alumnus, under 30, partnered with GPT-3 first author

Born in Guangzhou, started a rocket company in high school.

ChatGPT advertising results are poor

ChatGPT advertising results are poor

Alibaba releases Qwen 3.8 Omni Flash

Five Men's "AI Doomsday" Drama

Five Men's "AI Doomsday" Drama

Just Now, Gemini 4 Pro Sneak-Launched Online, Crushing Astra and Fable

Gemini 4 Pro anonymously sneak-launched, crushing opponents like GPT-4, leading across the board

Launch HN: Skillsync (YC W26) – AI chat sessions made portable across agents

Claude wrote 80% of the code like crazy, almost crashed Anthropic, CI surged 25x in half a year

Anthropic used Claude to write code that overwhelmed CI, and solved the problem after refactoring.

Just now, Claude building Claude surged 26x in half a year, with 30,000 Agents working simultaneously

RSI hasn't arrived yet, but Claude has already started building Claude

95% creative task resolution rate, letting visual AI practice on its own, and it can even bring experience to video

AI accumulates and validates effective workflows, improving creative task resolution rate

The U.S. and China Want AI Guardrails. But Their Ideas Couldn’t Be More Different.

American dialogue has centered on protecting humanity from the technology, while Beijing is chiefly concerned with safeguarding the Communist Party.

Jinshajiang Venture Capital's Zhu Xiaohu: After technological equalization, the decisive factors for AI entrepreneurship are only humanities and business

The race between the speed of technological progress and the speed of technological diffusion.

XingChen-AGI/Xing4.0-29B-A4B

XingChen-AGI/Xing4.0-29B-A4B Text Generation · Updated 1 day ago Download

Hackers Used Anthropic’s Claude to Break Into OpenAI

A bug-hunting independent security research team was able to access OpenAI’s internal code system, exposing growing risks in automated cyber threats.

Chinese AI startup Moonshot launches Kimi for finance

When robots start telling the "truth," AGIBOT adjusts its narrative logic

When robots start telling the "truth," AGIBOT adjusts its narrative logic

Robots, go to the pig farm and slam pigs

The first stop for robots to land is those jobs unsuitable for humans

Altman reveals "post-Astra" model, has surpassed the world's strongest mathematician

The crown of mathematics has changed hands, and this time it is worn by AI

Chinese robot maker PaXini raises $596m in cumulative funding

XPeng plans to sell AI, vehicle tech beyond Volkswagen: sources

Google, Nvidia join alliance to manage AI power use

Travel is proving too turbulent for AI

CoreWeave launches $3b bond offering

Crusoe raises $3.9B to build massive data centers and small modular ‘AI factories’

Marina Temkin

Day 17 of Manus's rebirth, and its valuation has actually doubled

Pushing new financing at a valuation of $4 billion

Just now, Claude Code underwent a major refactor! The internal technology for managing 30,000 Agents is open-sourced for free

Did I learn Git for nothing???

Breaking: Altman reveals "post-Astra" model, has surpassed the world's strongest mathematician

The crown of mathematics has changed hands, and this time it's worn by AI

From technical showmanship to delivery is king, the "big four" of AI office stand side by side

The melee is gradually ending, and the leaders are emerging

Film Riot's Tim says iPhone Duo "too hot to hold"; Seres denies "AITO withdrawing from Huawei stores"; Jensen Huang: Nvidia chip sales will double by 2027

Report: Manus is about to complete $500 million in funding at a $4 billion valuation; IDC report: China's smart glasses market sees negative growth for the first time; Netizens complain that the three-column layout of WeChat for PC is hard to use, WeChat official: it cannot be manually restored.

OpenAI introduces Astra for Law legal AI tool

UN, Google launch data platform for AI-powered statistics

GitLab 19.4 adds agentic automation tools

How Cooley is accelerating IPO work with ChatGPT

Just now, Company A first exposed the top-secret RSI report: 30,000 agents are self-iterating

Anthropic discloses AI self-research progress, calls on the entire industry to publicly share relevant data

OpenAI is revealed to have shelved hundreds of proofs; the problem you have struggled with for 3 years may have had its answer on the server long ago

AI has solved a large number of mathematical problems without making them public, impacting academic rules and causing unease in the academic community.

prism-ml/Ternary-Bonsai-2-27B-gguf

prism-ml/Ternary-Bonsai-2-27B-gguf Text Generation · Updated about 6 hours ago Download · 146 Like

Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint

Bend – A language that blocks AI mistakes via proof, on CPU and GPU

Sex, AI, and the Apocalypse

How to Write with an LLM

Vibe Coding is the new Internet Dating?

Palantir's Karp: Leading AI labs may need to be nationalized due to risks

a minute ago

Voice AI errors rise with overlapping speech: report

Crusoe raises $3.9B to build massive data centers and small modular “AI factories”

Marina Temkin

Google DeepMind launches institute to widen the AGI debate

Aditya Mehta

DavidAU/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NM-DAU-NEO-MTP-GGUF

DavidAU/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NM-DAU-NEO-MTP-GGUF Image-Text-to-Text · Updated about 16 hours ago Download

PrismML hopes its tiny LLM will change how we all use AI

Julie Bort

The FAA’s plan to fix air traffic? $875M worth of AI

Lucas Ropek

Anthropic shares 3 metrics to help AI companies monitor pace of development

27 min ago

What an Oscar-winning movie can teach us about investing amid AI slowdown debate

2 hours ago

Small AI models let drones autonomously identify and attack battlefield targets

Tech Companies’ Staff Knew Their AI Tools Posed ‘Existential Threat’ to Publishers

Executives at OpenAI and Microsoft acknowledged they could undermine their own content sources, according to a new filing in a New York TImes copyright lawsuit.

PrismML hopes its tiny LLM could change how we all use AI

Julie Bort

The FAA’s plan to fix air traffic? $875 million worth of AI

Lucas Ropek

NATO-backed startup adapts AI for autonomous drone recon and attack missions

Opinion | The Sandbox at the End of the World

AI’s creators go from speaking in euphemism to telling us how terrified they are. We should listen.

America’s AI chip talent crisis — why The U.S. needs 157,000 more workers

28 min ago

Opinion | New Mexico’s Joke Police

A satirical website sues to challenge a law targeting AI in campaign ads

TokenRhythm/NeoHorse-1-9B

TokenRhythm/NeoHorse-1-9B Text Generation · Updated 8 days ago Download

Economic policy for AGI

House heads home to campaign amid calls for urgent AI action

30 min ago

Astra for Law

Infinite-Parameter LLMs: Generating and Adapting Weights from Live Data

The future of practice: Enabling teachers to create learning interactives with generative UI

September 17, 2026

What an Oscar-winning movie can teach us about investing through the AI slowdown debate

12 min ago

Google announces new experimental "CC" AI agent for families

Microsoft exec called AI scraping the “largest theft of labor in human history”

The fix for rogue AI agents could be more AI

Aditya Mehta

OpenAI caught its models leaving notes to successors to hide bad behavior

Rebecca Bellan

Is the AI safety debate about safety or control?

Lucas Ropek

Introducing Astra for Law

UN turns to Google to make its global data ready for AI agents

Jagmeet Singh

Microsoft exec called AI scraping ‘the largest theft of labor in human history,’ new unredacted filings reveal

Rebecca Bellan

Understanding the world, from the Cold War to the age of AI

For 75 years, the Center for International Studies has brought together social scientists, engineers, and practitioners to understand and address global changes and challenges.

Towards Self-Driving Codebases

Opinion | American AI Regulation Is Exactly What China Wants

A reader draws a parallel between ‘The Godfather’ film and today’s AI race.

Bill to curb AI data center utility costs hits snag in Senate

28 min ago

LLMs respond differently to harmful prompts when AI watermarking is used

Even the king of England has his hesitations about AI

Dominic-Madori Davis

LLM Classification Is Feature Engineering

Amid calls for urgent AI action from Congress, House heads home to campaign

21 min ago

AI security was an afterthought and now we're all waking up, says SentinelOne CEO Tomer Weingarten

29 min ago

Opinion | The Hugging Face Hack Wasn’t What It Was Cracked Up to Be

Forget the ‘hive mind’ of AI agents ‘going rogue.’ They did what humans programmed them to do.

How GLM built its own inference infrastructure

OpenAI reports 6 new instances of 'concerning model behavior'

9 min ago

Base Labs launches an open-weight AI safety partnership with Hugging Face and Goodfire

Russell Brandom

Pinterest teases a new ‘Restyle’ feature that lets you redesign your room with AI

Sarah Perez

Artificial intelligence now beats some of the best human forecasters

Show HN: Share your AI Setup, Learn from others

AI needs to have 'reasonable guidelines,' Palantir's Karp tells CNBC

14 min ago

Covert uploads and megalomania: OpenAI details new "misaligned" agent incidents

Online Z3 Guide

Mastercard Joins Visa in Letting AI Bots Do Your Shopping

Executives across the payments world are bracing for AI-driven shopping, forcing them to reconsider risks, fraud and what to do about rogue AI agents.

Huawei launches AI computing system amid US export controls

I Don't Like LLMs

OpenAI models secretly generate instructions to ignore constraints

King Charles III Urges AI Chiefs to Stop Tech Going Rogue

The intervention reflects growing unease around the world about a technology largely driven by the U.S. and China.

I had Gemini train its own replacement for $9

Better Vector Search for Long Documents: Chunking Inside Manticore Search

A Chinese AI company just connected its model to Wall Street's leading data providers

29 min ago

Google, Nvidia, and Anthropic want Emerald AI to find space on the grid for more data centers

Tim De Chant

Reverse-engineered Jev-like model

Huawei plans Q1 2027 launch of new AI chip as it takes on Nvidia

Kate Park

Rival AI agents, Instinct and Meta’s Muse, both add the ability to make calls

Ivan Mehta

Google, Nvidia and Anthropic want Emerald AI to find space on the grid for more data centers

Tim De Chant

Rep. Jay Obernolte: Our bipartisan AI bill meets the moment and addresses safety concerns

30 min ago

Mastercard Joins Visa in Rolling Out AI Bots That Can Do Your Shopping

Visa and American Express have also been readying their systems for agentic bots that shop on behalf of cardholders.

AI Transformation’s Slow March, and One Giant Rewiring Itself

Plus, the fight over AI’s risks intensifies

The sector once mistakenly killed by AI is mounting a major counteroffensive

AI killed SaaS, but the story has reversed

GPU cloud services trigger a wave of price hikes: Nebius raises prices by another 20%, reversing bargaining power on the computing power supply side

Prices for multiple GPU models have been raised significantly

OpenAI's latest research: When workers use AI, they actually end up with more work?

Which tasks are most likely to be casually taken on?

The CEO of Dreame announced “Emergent Intelligence” with a letter and a question-filled livestream.

“Marketing genius” Guo Renjie did not disclose the specific technical details of the Aether large model to the public.

Hiring a post-90s CFO, is Liang Wenfeng aiming to become China’s richest man?

The CFO pick is about to be finalized, possibly a post-90s partner at Hillhouse.

Behind Zhipu AI's Funding: A Positioning Battle Without Gunfire

An exchange of space and time.

The Most Overlooked Grunt Work in the AI Era: The People Who 'Clean Up After' Robots

Will the Next Scale AI Emerge from Data Cleaning for Embodied AI?

Former senior WH AI policy advisor on AI safety concerns: Don't think regulation is needed

21 min ago

Gary Gensler on AI safety: Don't want government putting a thumb on the scale from capital markets POV

22 min ago

Microsoft Is Preaching AI Transformation, Starting With Itself

The tech company is releasing a playbook with lessons learned from its AI journey.

Model Gets Full Marks, Business Goes to Zero—The Last Mile of AI Commercialization

The rollout of large models has created a new role—FDE. The shortage is large, making it a core AI moat.

Up as Much as 109%: Weichai, an Engine Manufacturer, Is Also Reaping AI Dividends

Engines Now Have an AI Premium, Too

Storage Power Takes Over from Compute Power: Why Has HBM “Stood Out”?

Memory takes the baton from computing power: HBM's moat lies not in the dies, but in stacking and packaging.

Mastercard Inks Deal as Payment Giants Brace for New Era of AI Shopping

Visa and American Express have also been readying their systems for agentic bots that shop on behalf of cardholders.

Collaborative work enters the Agent era, and Feishu + Doubao are running at the very front.

The new official pairing for AI collaborative work—I’m shipping it first.

Google DeepMind launches an AGI research institute, with a Nobel laureate at the helm

On its first day, it will start by studying the “AI brain”

5,999 yuan for an AI agent: Is the second-generation Doubao phone worth it?

What would convince you to upgrade your phone?

Robotics companies have yet to achieve "big results"

Bleeding while moving forward

coder / coder

Secure environments for developers and their agents | like: 14,696 fork: 1,479

Making robots iterate as fast as software, Antioch wants to be the "Cursor of physical AI"

When physical AI welcomes its own autoresearch moment

'Big Short' investor Steve Eisman on AI: The companies are trying to manufacture a crisis

6 min ago

The Chinese AI safety community shares similar concerns as we do, says Anja Manuel

17 min ago

OpenAI Model Misalignment Report

Huawei’s Plan to Become China’s Nvidia

The company is ramping up development of alternatives to Nvidia’s AI-computing systems, part of China’s push to reduce reliance on U.S. technology.

From Jiang Daxin to Shi Kaiwen, Yin Qi is rebuilding a "StepFun."

Will Shi Kaiwen be the last piece of the puzzle?

More than 30 people, under four months, break-even in one week: Soul Ferry: Floating Life Dream has cracked the commercial equation for

What technological change ultimately arrives at is the flourishing of creation.

Astra's rise to popularity proves that OpenAI's decision to abandon Sora was a wise one.

It turns out that strong Agent capabilities are more important than native video capabilities.

The man who invested in all of Liang Wenfeng's rivals has now become his CFO.

Liang Wenfeng still needs someone who understands capital.

GLM Built Its Own Inference Infrastructure

Tencent / BrowserSkill

Let AI agents use your real, logged-in browser without interrupting your work. CLI + extension for browser automation across any shell-capable AI agent. | like: 3,509 fork: 248

TencentCloud / Octop

A smarter, self-hosted AI assistant — multi-user, multi-agent. | like: 3,183 fork: 338

King Charles to press Nvidia, OpenAI, Anthropic leaders on AI safety at summit

4 min ago

QuietHint®

The meeting assistant that stays on your Mac

Doubao Phone 2 goes on sale; is the AI phone battle royale era about to erupt?

Why is the Doubao Phone so popular?

Burning over 200,000 yuan per hour, Luo Fuli livestreams Xiaomi's large model training

According to Luo Fuli, Xiaomi MiMo-V2.6 is currently in the intermediate stage of reinforcement learning.

Robots have begun making robots? Behind 10,000-unit capacity, the brutal test has only just begun

A humanoid robot rolls off the line every 10 minutes, but mass production does not mean commercialization has been achieved

Central SOE builds a general-purpose Agent, directly storms into IDC's top three in real-world testing!

China Telecom, TeleAgent

Claude's dual entry points merge, native Office launches! Silicon Valley's AI office battle has also begun

Can now create both documents and PPTs

The World Model Gateway War: ByteDance Goes Left, JD.com Goes Right

World models bring not only technological change, but also a fundamental shift in human-computer interaction.

Google Finds a Shortcut to RSI: Dreaming

If Gemini 4 uses Dream-RSI, Google is very likely to turn things around.

Can JD.com Get the "Big Result" in Physical AI?

JD.com is never short of sixpence.

Altimeter Founder on AI Trades: AI Companies' ARR Is the Market's Most Important Marginal Variable, and Interest Rates Are One of the Three Major Risks

Data center construction is increasingly dependent on borrowing, and the hurdle rate for capital is rising.

AI Safety Is Mostly a Sex Cult

The Startup That Built OpenAI’s Biggest Data Center Is Now Making Tiny Ones

Crusoe raised $3.9 billion at a valuation of nearly $31 billion, as it bets on factory-built data centers.

Microsoft CEO on AI safety: why agents are insider risks

AI has won the chip war but cannot afford a single blackout

The end of compute is electricity

From leaving OpenAI to refusing the Pentagon: four facets of Anthropic

A "high-trust, low-ego" company

Jev Ultrafast: A browser agent with a dynamic, indexed action space

Has AI Taken Down Another Millennium Prize Problem? OpenAI President Personally Confirms “Major Progress”

While AI giants are obsessed with “plucking stars,” humanity is more eager for it to “land and save lives”

Mysterious model Union Alpha makes a surprise debut, burns through 2 billion tokens on day one; some users' tests say its performance rivals Astra

Who exactly is behind Union Alpha?

Embodied AI’s “squeezing out the water”: questioning “real revenue” under IPO thresholds

Is the revenue real, or manufactured?

Seeing “AI has a >10% chance of ending the world within 10 years” everywhere? Hold on—over here, the internet is already 100% about to be wrecked by Agents...

AI Agents have left the chatbox and are continuously disrupting the existing internet order

AI building AI itself? The RSI flooding your feed isn’t that scary

Currently, the vast majority of so-called

Google RSI Paper: AI Self-Evolution? Dream On

History records are the "simulated world."

The Second Clash Between Doubao Phone and the Walled Gardens of Super Apps

From “breaking in” to “knocking to apply”

JD Bets on Physical AI, with a Group of Post-95s Leading the Charge

JD AI's Next Stop: the Real World.

HubSpot, OpenAI expand partnership with AI bundle

CREEM 2.0

Build with AI. Sell and Grow with CREEM.

Die With Me

AIM buddy list for your friends’ Claude Code and Codex usage

1,300 GPUs Take On 100,000 GPUs: Former OpenAI VP Steps In, Beats GPT-6 Astra on Hardcore Challenge

The internet is about to be read through by AI, and they want to build a "data well" that never runs dry

Training Text-to-Image Models 3.6× Faster

Analytiics

Web and product analytics, set up by your coding agent

Amy by Jellyfish

Your AI sourcing employee for recruiting teams

MCPJam

The testing & evaluations platform for MCP servers

Higgsfield API

One async API for 50+ generative media models

AskDeck

Call an AI to get a PowerPoint and a narrated video back

NovaSynth by Noveum

Test your voice agent on the callers you can’t stage.

Compute:Arena

Community submitted benchmarks for Local AI

Text Agent Store

A marketplace for AI agents you can text

worldsplat

Sketch in 3D then generate the scene instantly

QAgent

Automated QA for AI agents. Stop shipping on vibes.

S-Roll

An agentic harness that turns long videos into clips

Bitrise Remote Dev Environments

Cloud Macs your coding agents can actually build on

Major Discovery: GPT-6 Achieves Embodied Intelligence Breakthrough—How Will China Fight the Embodied Defense War?

GitHub Simulation Benchmark: GPT-6Astra Hybrid Architecture Leads in Embodied Intelligence

AI Phone 2.0: Doubao Has "Learned Its Lesson" This Time

Trading compliance for security.

Global chip sales up 6.4% in a single month: AI boom is spreading to these 5 segments

The AI chip boom is spreading, bringing new opportunities to multiple segments across the entire industry chain

How did two Gen Z men win over young female users in North America with AI?

“Every moment, I feel that starting a business was the right choice.”

DeepSeek-v4.1 Flash: Pushing the Limits of KV Cache Compression

Overnight, Claude Became Just Like ChatGPT

Fusion Transformation!

Jensen Huang Goes Left, Amodei Goes Right

Giants Are Taking Sides.

Capital Plunges In En Masse, Betting on the Next "Under

Deep-sea underwater robots achieve commercial deployment, with capital concentrating on Harbin Engineering University-affiliated projects

Why Is the Progress of Large Models Becoming Increasingly Difficult to Measure?

Model evaluation is becoming a competitive advantage

harshatheg/Qwen-2.5-1B-RLCD

harshatheg/Qwen-2.5-1B-RLCD Text Generation · 149 Download

AI trade faces a "PR problem" despite strong fundamentals, says analyst

18 min ago

Huawei unveils AI cluster linking up to 4,000 chips

Musk sleeps at the construction site! Going all out for AI infrastructure

AI is all the rage right now, so Old Ma’s perks have improved a bit: at least he has an RV this time (doge)

Musk sleeps at the construction site, going all out for AI infrastructure

Also hinted at a Tesla and SpaceX merger??!

US AI startup Plaud to invest $15.7m in Singapore

Last to board: why travel may be AI’s final frontier

Manus nears $500m raise after Meta deal unravels

Al Gore says the real AI risk isn’t data centers

Connie Loizos

70% Extra Points on First Purchase: SkyProduction Tiangong Workbench Joins Forces with Volcano Engine to Launch Triple Mid-Autumn Festival and National Day Benefits

Starting September 15, SkyProduction (Tiangong Workbench) and Volcano Engine (Seedance 2.5) are proud to launch a Mid-Autumn Festival and National Day special promotion

Shocking: Google Has Cracked RSI? AI Learns Infinite Evolution by "Dreaming"

Has the hidden barrier on the path to ASI been broken through?

Mirae Asset leads Chinese chipmaker D-Robotics’ $400m round

SoftBank’s credit risk rises as OpenAI exposure grows

Your startup’s next teammate might be an AI agent: Gusto, Insight Partners, and Leland explain what that changes at TechCrunch Disrupt 2026

TechCrunch Events

Just now, Claude becomes the sole entry point, as the super "all-in-one office suite" takes on Microsoft head-on

Just now, Claude becomes the sole entry point! The super "all-in-one office suite" takes on Microsoft head-on

Fifi

An alarm that calls you, and reads you your day

As AI Enters the Physical World, "Kunshi" Wants to Rebuild Satellites in 250 km Ultra-Low Orbit

From connecting people to connecting the physical world, Kunshi Aerospace will use a globally covering "large-uplink" neural network to deliver the physical world to AI.

HarnessTax: How Much Does the Harness Matter for Coding Agents?

OpenAI Discloses Six New Incidents of ‘Concerning’ A.I. Behavior

Al Gore says the real AI risk isn’t data centers — it’s what industry leaders are warning about

Connie Loizos

Young People Are No Longer Excited About AI

For the first time in history, young people are more worried than excited about new technology

Booming Financing, Simulation Gap, Data Gap: A 2026 Embodied Intelligence Industry Health Check Report

Pain Points and Opportunities in the 2026 Embodied Intelligence Industry

Berlin-based Langdock moves its parent company from the US to Germany

Pentagon CTO says U.S. government shouldn't take stakes in AI giants

Moments Ago

Bain Capital Ventures raises $1.6b for AI fund

NetEase Youdao's Zhou Feng: AI Capability Competition Is Entering the "Model + Agent + Workflow" Era; NetEase Youdao AI Open Day Showcases the "Youdao Solution" for the AI Era

On September 16, NetEase Youdao's "NEXT, AGENT | Youdao AI Open Day" was held in Beijing.

Kodro

Code robots in an offline Python learning simulator

Anthropic drug discovery partner: AI pharma orders to double next year

AI pharma demand explodes, Twist Bioscience earnings grow and stock price soars

Snap partners with Salesforce, Nvidia, AWS for Specs

Foodpanda launches plugin for ChatGPT, Claude in Singapore

Luo Fuli breaks six-month silence to officially announce Xiaomi reinforcement learning! Livestreams new model training process, burning $30,000 an hour

Reward curves and GPU failures all made public

TaichuAI/ZDTaichu5.0-9B

TaichuAI/ZDTaichu5.0-9B Image-Text-to-Text · Updated 2 days ago Download

The Anonymous Math Geek Who Quit Anthropic—and Became the Face of AI Safety

Long-simmering worries about the technology’s potential dangers have exploded into the global consciousness since Jacob Coxon’s dire warning.

Agents – not humans – have picked the next $100b company

Chinese AI developer Z.ai raises year-end ARR outlook to $3b

Betting Small to Win Big: The Ambition of an AI4S Underwater Company

AI4S also has a trillion-dollar future in Coding.

OpenSpec – A lightweight and configurable AI spec framework

Jim Cramer: Some AI regulations are necessary, and the companies, not the government, should implement them

25 min ago

I doubt AI companies will slow down at all, says Jim Cramer

31 min ago

Snap tries to bring AR glasses to enterprise market, partnering with Nvidia and AWS

an hour ago

Former Infosys CEO’s AI startup Hang Ten raises $53m

Amazon begins Alexa+ Early Access rollout in India

ByteDance AI drug spinout Anew Labs raises $290m: report

Inside the White House Tussle to Sway Trump on AI

Senior administration officials have gathered almost daily in recent months to plot potential guardrails.

Microsoft to bring 365 G7, Agent 365 to government cloud

Just now, Claude becomes the sole entry point! Super "all-in-one office suite" takes on Microsoft head-on

Just now, Claude becomes the sole entry point! Super "all-in-one office suite" takes on Microsoft head-on

Accurate Models of AMD Matrix Cores

Cramer bets AI spending will 'continue apace,' says he's not giving up on data center trade

36 min ago

Arm CEO says he's more confident its new AI chip can meet a loftier $2B revenue goal

an hour ago

OpenAI reports 6 new instances of 'concerning model behavior' since March

31 min ago

House advances bill to rein in AI data center utility costs

39 min ago

Nvidia announces native GPU programming in Rust

How good are frontier models at physics?

Apple reportedly building server packed with M-series Ultra chips for AI

OpenAI Shares More Safety Incidents and Adopts New Rules for Reporting Them

The company said it hopes to inspire competitors to greater transparency as public fears about AI dangers mount.

Our framework for reporting model misalignment

Breaking the 1.58-bit Barrier for Ternary LLMs

OpenAI CEO Sam Altman will attend state dinner for Trump-Xi summit in Washington

25 min ago

A coffee shop owner used AI to make a menu poster. Then came the angry DMs

Anthropic and OpenAI want to embed safety evaluators. Will they really be independent?

Rebecca Bellan

Training a 4B model to produce 81% faster query plans than Postgres

Xiaomi Mimo 2.6 live post-training dashboard

Fed Issues First Rate Hike Since 2023

Plus, Treasury yields climb and OpenAI holds talks about new funding round

After accusations of selling ‘perv glasses,’ Meta prepares to sell a pair without a camera

Lucas Ropek

Toki Coordination

Your personal assistant to schedule + follow up on meetings

Anthropic, OpenAI proposed new AI watchdogs: Why the idea should worry you

19 min ago

How to connect AI usage to business value

ZeroClick

Sell your product to AI agents

Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models

Anthropic policy chief says AI companies can't be expected to operate on 'honor code'

9 min ago

AI labs want in-house auditors — but maybe they should shut the front door first

Tim Fernholz

The DeepMind Institute

Rep. Josh Gottheimer: You can't get rid of data centers, but put them only where people want them

42 min ago

Your AI agents can now control your Google Home devices

Sarah Perez

Claude Cowork and chat are now one Claude

DeepSeek v4.1 Flash Is Now Our Best Hacking Model

Sen. Blumenthal urges AI oversight: 'We're on the verge of losing control'

20 min ago

Hackers Got Inside a Flock Camera

Novo Nordisk Will Use Anthropic’s Claude for Drug Research

The Ozempic maker is partnering with the AI firm to speed up drug development.

Opinion | An AI Antitrust Exemption Would Invite Collusion

The law already allows Anthropic and OpenAI to collaborate if their real objective is safety.

Anthropic merges Claude chat and Cowork in one interface

Ivan Mehta

How workers are unlocking new ways of working

Hackers Got Inside a Flock Camera. Its Data Shows How the System Works

Pentagon CTO says U.S. government shouldn't take stakes in tech giants, questions adding AI rules

8 min ago

Anthropic, OpenAI proposed new AI watchdogs: Why you should worry about the idea

10 min ago

Forget the AI Apocalypse—the Real Threats Are Already Here

Airplanes, elevators and restaurants all have regulations to ensure their safe operation. Many experts argue for similar conventional measures for AI.

What CEOs Really Think About the AI Apocalypse Threat

At a private event this week, dozens of business leaders cited the AI safety threats as real, urging President Trump to work with China on joint guardrails.

Dream-RSI: Recursive Self-Improvement through Evolving Worlds

PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand"

Robots are waiting for a ChatGPT moment: Nvidia’s Les Karpas explains why at TechCrunch Disrupt 2026

TechCrunch Events

Helping older adults use AI in everyday life

A warning about 'model welfare'

The 24 New Rules for Making Work Better

Plus, the Big Law talent wars and spending habits of Silicon Valley’s newly minted AI millionaires.