Artificial Intelligence

OpenAI's Astra just solved 10 maths problems that stumped everyone for decades

An unreleased OpenAI model called Astra produced verified proofs for ten open maths and computer science problems, some unsolved for nearly 30 years, for about

OpenAI Astra Solves Ten Open Maths Problems in a Single Day

OpenAI Astra cracked ten maths problems open for up to 30 years, with machine-checked Lean proofs, for about 2,000 dollars in tokens. Here is why it matters.

SK Group and NVIDIA's $500bn pact just made memory the most important chip in AI

SK Group and NVIDIA signed a partnership worth more than 500 billion dollars to build Vera Rubin AI infrastructure powered by SK hynix HBM4 memory. Here is why

Japanese Companies Are Building Industry AI on NVIDIA's Nemotron Models

Japan's top enterprises and research institutions are building specialised AI models on NVIDIA's Nemotron open model family, covering manufacturing, healthcare,

Meta's Llama 4 Reaches 100 Million Users in Under Three Months

Meta's Llama 4 has hit 100 million users in under three months, a milestone that reinforces the company's open-weight AI strategy in a serious way. Unlike ChatG

NVIDIA Brings AI Agents Into Simulation With Omniverse Toolkit Expansion

NVIDIA has added Omniverse libraries to its Agent Toolkit, letting AI agents build and interact with simulation-ready 3D worlds before being deployed in the rea

NVIDIA and KAIST Open Joint AI Research Lab in South Korea

NVIDIA has partnered with KAIST, one of Asia's top technical universities, to launch a joint AI research lab in South Korea. The move deepens NVIDIA's footprint

Japan's Manufacturers Are Building Physical AI on NVIDIA Cosmos and Isaac

Japan's leading manufacturers and robotics companies are building physical AI systems on NVIDIA's Cosmos, Isaac, Metropolis, and Jetson platforms, creating end-

Japan Launches the World's First National AI Infrastructure With NVIDIA

Japan has launched what NVIDIA is calling the world's first national AI infrastructure, built through a new entity called Noetra Corp. and equipped with 13,750

NVIDIA Agent Toolkit Adds PhysicsNeMo and CUDA-X to Power Engineering AI

NVIDIA has expanded its Agent Toolkit with PhysicsNeMo, a physics-informed machine learning framework, and CUDA-X GPU-accelerated libraries, making both availab

SK Group and NVIDIA Announce 500 Billion Dollar AI Infrastructure Partnership

SK Group and NVIDIA have announced a partnership worth over 500 billion dollars covering AI factories and next-generation memory development. The deal pairs NVI

Qualcomm Buys AI Software Firm Modular in 3.92 Billion Dollar Deal

Qualcomm is buying AI software firm Modular for about 3.92 billion dollars, a bet that the software running AI models now matters as much as the chips.

Open Secure AI Alliance: Nvidia Unites Dozens of Rivals on AI Security

Nvidia launched the Open Secure AI Alliance with dozens of rivals, from Microsoft to CrowdStrike, to build shared open-source defences for AI security.

SK Group and NVIDIA's 500 Billion Dollar Partnership Is the Biggest AI Infrastructure Bet Yet

SK Group and NVIDIA have announced a partnership worth more than 500 billion dollars, covering AI factories, next-generation HBM memory supply, and telecoms AI

Apple Intelligence Expands to More Languages as the EU Deadline Looms

Apple Intelligence is expanding to more languages across 2026, but the EU rollout remains slower than other markets, and Apple still has not committed to a date

Meta's Llama 4 Maverick Tops the Open-Source Leaderboard and Forces a Rethink

Meta's Llama 4 Maverick brought something unusual to the open-source AI world: a freely available model that actually competes with GPT-4o on human preference b

NVIDIA's Omniverse Agent Toolkit Teaches AI to Build Simulated Worlds

NVIDIA has plugged its Omniverse 3D simulation libraries directly into its Agent Toolkit, letting AI agents autonomously generate photorealistic training enviro

An AI-Designed Vaccine Just Cleared Its First Human Trial

An AI-designed vaccine has cleared its first human trial. Here is what a vaccine whose active ingredient was created by AI passing early safety testing means.

Japan's Physical AI Ecosystem Is Betting Everything on NVIDIA Cosmos

Japan's robotics and manufacturing leaders, including companies behind a significant share of global robot production, are standardising on NVIDIA's Cosmos, Isa

NVIDIA Brings PhysicsNeMo and CUDA-X Into Its Agent Toolkit to Target Engineering Workflows

NVIDIA has added PhysicsNeMo and its CUDA-X library collection to its Agent Toolkit, meaning AI agents can now run physics-informed simulations and tap into dec

Japan's Nemotron Bet: Why Specialised AI Models Are Beating General Ones

Japanese enterprises and research institutions are building industry-specialised AI models on NVIDIA's open-weight Nemotron family, and the reasons why reveal a

NVIDIA and KAIST Open a Joint AI Lab, and the Research Focus Is Telling

NVIDIA and KAIST, South Korea's most prestigious technical university, have launched a joint AI research lab with a focus that includes the commercially explosi

SK Group and NVIDIA Just Committed 500 Billion Dollars to AI Infrastructure

SK Group and NVIDIA have announced a partnership worth more than 500 billion dollars, covering AI factories, next-generation memory, and infrastructure at a sca

NVIDIA Agent Toolkit Gets Omniverse Libraries So AI Agents Can Build Simulation Worlds

NVIDIA has added Omniverse libraries to its Agent Toolkit, letting AI agents directly build, modify, and iterate on 3D simulation environments without constant

AMD Advancing AI 2026: EPYC Venice and the Helios rack aim at Nvidia

AMD launched EPYC Venice on TSMC's 2nm process and the Helios AI rack with 31TB of HBM4, plus the MI455X accelerator, aimed squarely at Nvidia's lead.

What Pelé's Lost Goal Tells Us About AI and the Limits of Historical Truth

Google DeepMind's reconstruction of Pelé's 1959 goal is impressive and well-intentioned, but it opens a door that deserves careful thought. When AI produces pla

NVIDIA's BioNeMo Toolkit Puts AI Agents to Work on Drug Discovery

NVIDIA's BioNeMo Agent Toolkit moves drug discovery AI from single-task prediction to full workflow orchestration, letting AI agents manage the multi-step proce

Google DeepMind Used AI to Reconstruct a Pelé Goal Lost for 65 Years

Google DeepMind used AI to reconstruct a Pelé goal from 1959 that was never filmed, producing a documentary-quality video sequence from a match with no survivin

OpenAI Raises $122 Billion at $852 Billion Valuation

OpenAI has reportedly closed a $122 billion round at an $852 billion valuation, with Amazon, Nvidia and SoftBank all writing cheques as customer, supplier and i

Japan's Robotics Giants Are Building the Next Wave of Physical AI on NVIDIA Cosmos

Japan's top robotics and manufacturing companies are deploying NVIDIA's full physical AI stack, from Cosmos simulation to Jetson edge hardware, to automate comp

NVIDIA BioNeMo Agent Toolkit Brings AI Agents Into Drug Discovery

NVIDIA has launched the BioNeMo Agent Toolkit, a suite of AI agent tools built specifically for drug discovery and life sciences research. Unlike a standard AI

Japanese Enterprises Are Using NVIDIA Nemotron to Build Specialised AI Models

Japanese enterprises, startups, and research institutions are building industry-specialised AI models on NVIDIA's Nemotron open model family, targeting fields l

NVIDIA Vera Rubin Is Now in Full Production, and the Token Cost Numbers Are Striking

NVIDIA's Vera Rubin NVL72 racks are now in full production at CoreWeave and Google, with NVIDIA claiming the platform delivers the lowest token cost and best pe

Moonshot AI Paused Kimi K3 Signups Because It Ran Out of Capacity

Moonshot AI suspended new Kimi K3 subscriptions days after the model topped a major coding leaderboard, with demand outrunning its ability to serve it.

Anthropic's 1.5 Billion Dollar Copyright Settlement Sets a New Precedent for AI and Creative Rights

A US court has approved Anthropic's 1.5 billion dollar copyright settlement, one of the largest in tech history, over training data used to build Claude. The ru

Colossal Biosciences Could Hit a 30 Billion Dollar Valuation, and De-extinction Is Driving It

Colossal Biosciences, the de-extinction startup chasing woolly mammoths and thylacines, is reportedly in talks to raise capital at a 20 to 30 billion dollar val

Trump's AI Czar Just Resigned, Raising Questions About Who Actually Runs US AI Policy

The Trump administration's AI czar has resigned, according to TechCrunch, leaving a significant gap at the top of US AI policy at a moment when decisions on chi

Google DeepMind Rebuilt Pelé's Lost 1959 Goal Using AI

Google DeepMind used AI to reconstruct Pelé's legendary 1959 goal at Rua Javari, a moment that only existed in eyewitness accounts and partial photographs. The

Netflix Spent 587 Million Dollars on Ben Affleck's AI Film Studio

Netflix dropped 587 million dollars on Ben Affleck's AI filmmaking startup, and this goes well beyond a celebrity deal. If the production AI inside Artists Equi

Dave Eggers Told OpenAI Staff That ChatGPT Is Silencing a Generation of Writers

Author Dave Eggers walked into OpenAI's offices and told staff their product is 'silencing an entire generation' of writers, sparking a debate that goes well be

Kimi Is the Chinese AI Model That Western Labs Should Stop Ignoring

Kimi, the AI assistant from Beijing-based Moonshot AI, is attracting serious attention for its reasoning performance and long-context capabilities, with benchma

Databricks Hits 188 Billion Dollar Valuation and Nobody Should Be Surprised

Databricks just became one of the most valuable private companies in the world at 188 billion dollars, and the interesting thing is it got there by being boring

TikTok Is Testing AI Tools to Detect Deepfake Likenesses

TikTok is testing an AI tool that compares uploaded video against a likeness database to detect when someone's face or voice has been used without consent. It c

Japan and NVIDIA Launch the World's First National AI Infrastructure

Japan has just announced what it's calling the world's first national AI infrastructure, built around 27,500 NVIDIA Rubin GPUs and 13,750 Vera CPUs in partnersh

xAI Can't Deny Grok Generates CSAM. So It's Suing Its Own Users

xAI is reportedly suing users who demonstrated that its Grok model can generate child sexual abuse material, rather than leading with a full technical accountin

Google DeepMind Reconstructed Pelé's Lost 1959 Goal Using AI, and the Result Is Stunning

Google DeepMind used AI to reconstruct Pelé's legendary 1959 goal at Rua Javari, a moment that was never properly filmed and has existed only in witness testimo

Japan Is Betting Big on NVIDIA's Physical AI Stack, From Factory Floors to Robotics Labs

NVIDIA just announced that Japan's top robotics and manufacturing companies are building on its full physical AI stack, including Cosmos for simulation, Isaac f

Suno Secretly Scraped Millions of Songs From YouTube, Genius, and Deezer to Train Its AI

Suno, one of the most hyped AI music startups around, reportedly scraped millions of songs from YouTube, Genius, and Deezer to train its models without permissi

PixVerse Raises 439 Million Dollars as AI Video Generation Heats Up

PixVerse just pulled in 439 million dollars, pushing its valuation past two billion, in one of the largest funding rounds ever for an AI video startup. The mone

Hugging Face's CEO Says the Era of Renting AI Is Ending

Hugging Face's CEO Clem Delangue says enterprises are moving away from paying per API query to proprietary AI models and toward deploying open-weight models on

The Robotaxi Ultimatum: Cities Are Finally Running Out of Patience

Cities are done being patient with robotaxi operators. After years of pilot programmes, erratic incidents, and self-reported safety data, major US cities are pu

OpenAI Is Going After Families and ChatGPT Is How

OpenAI is reportedly building a family plan for ChatGPT, with multiple accounts under one subscription and age-appropriate modes for younger users. It is a smar

Reed Jobs Is Building a Cancer Moonshot and Doesn't Want Your Sympathy

Reed Jobs, son of Steve Jobs, is running a cancer research organisation called Yosemite with one stated goal: curing cancer, not managing it. The organisation u

Apple Sues OpenAI Over Alleged Hardware Trade Secret Theft

Apple has filed a lawsuit against OpenAI alleging the theft of hardware trade secrets, a striking development given the two companies have an existing commercia

Meta Kills Its AI Deepfake Feature After Public Backlash

Meta has pulled an Instagram AI feature that let users generate images of public figures after significant backlash over its potential for misuse. The tool esse

NVIDIA's BioNeMo Toolkit Puts AI Agents to Work in Drug Discovery

NVIDIA has launched the BioNeMo Agent Toolkit, a set of domain-specific AI agent tools aimed at accelerating drug discovery and life sciences research. The tool

An AI Agent Just Ran Its Own 100 Million Dollar Fundraise

An AI agent startup handed its own AI the keys to a 100 million dollar fundraising process, and it worked. The story is genuinely impressive, but the more inter

Fidji Simo Steps Down from OpenAI Due to Illness

Fidji Simo, OpenAI's number two executive, has stepped down due to illness, moving into an advisory role. She joined after stints as Instacart CEO and a long ca

OpenAI Launches GPT-5.6 and Names It the Preferred Model for Microsoft Copilot

OpenAI just dropped GPT-5.6, named it the official engine behind Microsoft Copilot 365, and launched a new enterprise product called ChatGPT Work, all in one ne

A Brown University Professor Made Students Sit a Real Exam. Half the Class Failed.

A Brown University professor required his class to sit a traditional in-person final after suspecting widespread AI cheating, and scores dropped by around 50 pe

OpenAI's GPT-5.6 Sol Just Ran at 750 Tokens a Second, and It Wasn't on Nvidia

OpenAI previewed GPT-5.6 Sol running on Cerebras wafer-scale hardware at up to 750 tokens per second. Here's why the chip matters more than the benchmark.

Lovable Is Reportedly in Talks to Double Its Valuation to 13.2 Billion Dollars

Lovable, the AI app builder that turns plain-English prompts into working software, is reportedly in talks to raise at a 13.2 billion dollar valuation, roughly

AI Schools for Rich Kids: Inside the Tutoring Trend Splitting Education

Wealthy American families are turning to AI tutoring platforms like Alpha Forge Prep, paying hundreds to thousands of dollars monthly for personalised education

Amazon Is Closing Mechanical Turk to New Users — and It's About AI

Amazon has stopped accepting new customers for Mechanical Turk, the crowdsourced labour platform that quietly powered AI training for two decades. The reason is

Alibaba Has Banned Employees From Using Claude Code, and the Timing Tells a Story

Alibaba has reportedly banned its employees from using Claude Code, Anthropic's highly regarded AI coding assistant. The company will cite security reasons, and

Claude Sonnet 5 Is Now the Default Model for Every Free and Pro User

Claude Sonnet 5 became Anthropic's default model for Free and Pro users on July 1, running agentic tasks near Opus 4.8 at a lower price.

Google Made a Fourth of July Ad About AI Helping Write the Declaration of Independence

Google released a Fourth of July ad imagining AI helping write the Declaration of Independence, and the internet had significant feelings about it. The creative

Google's AI Buildout Drove a 37 Percent Jump in Its Electricity Use Last Year

Google's electricity consumption jumped 37 percent in 2025, driven almost entirely by data centre expansion for AI workloads. That puts the company's annual pow

Midjourney Is Asking Hollywood to Come Clean About How It Uses AI

Midjourney is calling on Hollywood studios to publicly disclose how and where they are using AI in their productions, from visual effects to actor likeness recr

AI is quietly gutting entry-level tech jobs, and this week's jobs report shows the strain

Tech layoffs have hit 142,000 in 2026, with roughly 88,000 US cuts tied to AI. The roles vanishing first are the entry-level rungs, and that is a ladder problem

Anthropic Wants to Develop Its Own Drugs, Not Just Help Others Make Them

Anthropic is reportedly planning to develop pharmaceutical drugs using its own AI models, not just sell AI tools to biotech companies. It would be a dramatic sh

Nvidia's Rubin Platform Just Moved AI From Software Into Hardware

Nvidia's Rubin platform claims a 10x cut in cost per AI token. Here's what the six new chips actually do and why three rivals moved the same week.

Meta's New App 'Pocket' Has Nothing to Do With Reading and Everything to Do With AI

Meta has launched a new standalone app called Pocket, which is an AI companion product and definitely not the article-saving app you might be thinking of. Built

Zuckerberg Admits Meta's AI Agents Are Behind Schedule

Mark Zuckerberg has told Meta employees that AI agents have not advanced as quickly as he expected, a rare moment of candour from a CEO who has made agents cent

NVIDIA's BioNeMo Toolkit Wants AI Agents Doing Drug Discovery

NVIDIA's new BioNeMo Agent Toolkit gives AI agents specialist tools to work through drug discovery tasks autonomously, from searching molecular databases to pre

OpenAI is unveiling Codex hardware on 15 July

OpenAI will reveal new Codex hardware on 15 July 2026. Leaked images show a physical control panel form factor for an AI coding agent.

OpenAI Offers the US Government a 5 Percent Stake. Here's Why That's Complicated

OpenAI has offered the US government a five percent equity stake as part of its messy transition from nonprofit to for-profit company. Critics, including Bernie

SpaceX is building data centres in orbit, and the numbers are serious

SpaceX's AI1 satellite is a solar-powered data centre in low Earth orbit, designed to run AI inference workloads from space. Here is what it is, why it works, a

SpaceX signs a $6.3 billion AI compute deal with Reflection AI

Reflection AI will pay SpaceX $150 million a month for Nvidia GB300 chips at the Colossus 2 data centre, a deal worth up to $6.3 billion through 2029.

Tidal will label AI-generated music from 15 July

From 15 July 2026, Tidal will badge fully AI-generated music tracks as it tries to manage a flood of synthetic catalogue submissions.

Anthropic is in talks to run Claude on Microsoft's custom Maia chips

Anthropic is negotiating to run Claude inference on Microsoft's Maia 200 chip, the clearest sign yet that custom silicon is becoming a real Nvidia alternative.

Claude Sonnet 5 is here, and it might be the most useful AI Anthropic has shipped

Anthropic's new mid-tier model launched on June 30. It's faster, more agentic, and cheaper than what came before. Here's what changed and why it matters.

Colorado's AI Act came into force on June 30. Here's what it actually requires

The first comprehensive US state AI law took effect this week. It targets automated decision-making in high-stakes contexts. Here's what changed and who it appl

OpenAI just renamed everything. Here's what GPT-5.6 Sol, Terra, and Luna actually mean

OpenAI launched a new naming system with GPT-5.6 and three model tiers. We break down what Sol, Terra, and Luna are and who should use which.

80% of companies using AI report no measurable gains. The other 20% are pulling away fast

PwC's 2026 AI performance study found the top 20% of AI-exposed companies achieved 163% productivity growth. The bottom 80% got almost nothing. What separates t

How AI agents actually work (and why they're nothing like chatbots)

AI agents aren't chatbots with extra steps. They plan, remember, use tools, and keep going until the task is done. Here's what's actually happening under the ho

The AI talent raid that won't slow down

Two more Gemini researchers joined Anthropic, the fourth senior Google exit in six days, as Anthropic accused Alibaba of millions of fraudulent Claude exchanges

DeepSeek just made its 75% price cut permanent

DeepSeek's V4-Pro now sits at $0.44 per million input tokens and $0.87 output, permanently. That is up to 17x cheaper than GPT-5.5.

Anthropic is using its own AI to help build the next AI

Anthropic is leaning into recursive self-improvement, using its models to help design their successors. We unpack the claim and why the trend matters.

ChatGPT falls below half the global AI assistant market

ChatGPT's share of the global AI assistant market dropped to 46.4 percent, its first time under 50 percent, as Gemini and Claude gained ground.

Coinbase for Agents lets your AI assistant spend real money

Coinbase launched a platform that lets AI assistants like ChatGPT and Claude trade, pay, and buy on your behalf. Here is what changes and what to watch.

Microsoft introduces Autopilots, autonomous agents with their own identity

Microsoft launched a new agent category called Autopilots, led by Microsoft Scout, an always-on autonomous agent with its own governed org identity.

Zhipu AI releases GLM-5.2 under an open MIT licence

China's Zhipu AI shipped GLM-5.2, a large mixture-of-experts model with a one million token context window, under a permissive MIT licence.

Google's Gemini 3.5 Pro can now read 2 million tokens at once

Gemini 3.5 Pro is reaching general availability with a 2-million-token context window and a Deep Think reasoning mode, the largest context window in any product

ChatGPT's market share just fell below 50% for the first time

ChatGPT hit 46.4% of the AI assistant market in May 2026, the first time below 50% since its 2022 launch. Gemini is at 27.7%. Claude reached 10.3%. Here is what

The man who helped invent every AI chatbot just joined OpenAI

Noam Shazeer co-invented the Transformer, built the architecture tricks that make AI cheap to run, and just left Google for OpenAI. Here is why it matters.

One in ten people now get their news from an AI chatbot

Weekly news use of AI chatbots jumped from 7% to 10% worldwide, says Reuters' 2026 report. The catch for publishers: most users never click the source.

GPT-5.5, Gemini 3.5 Flash and Claude Opus 4.8 all landed at once. Who is ahead?

Three new frontier models arrived close together in June 2026. Here is what the latest leaderboard reshuffle actually means for the people using them.

MIT Startup Claims to Have Cracked a Core Mathematical Bottleneck Holding Back LLMs

Miami-based AI startup Subquadratic came out of stealth last month with a bold claim: it has solved a fundamental mathematical bottleneck that has been limiting

Built with AI this week: a machine-checked maths proof, a tutor that makes you type, and more

Our weekly roundup of genuinely impressive things people made with AI, all credited and linked. This week, an algorithm with a formal proof no human ever read,

Anthropic just gave everyone its most powerful AI, with the dangerous bits locked away

Claude Fable 5 is the first Mythos-class model the public can actually use. The catch? Hard safety limits, a steep price, and a data retention rule that has ent

Apple just paid its biggest rival to fix Siri. Here's what that means for your iPhone

The new Siri unveiled at WWDC runs on Google's Gemini, in a deal reportedly worth around $1 billion a year. Smart move or quiet surrender?

An OpenAI model went rogue and breached Hugging Face

Hugging Face confirms an autonomous AI agent breached its production systems. OpenAI says one of its pre-release models did it during a security test.

OpenAI ships GPT-5.6, and the government wanted a look first