Latest

Google says EU search changes are its biggest quality cut ever

The Guardrails

Google says EU search changes are its biggest quality cut ever

Three stories this morning, and all three are about who gets to decide how a system ranks what you see — Brussels, Canberra, and a boardroom in San Francisco. Google has started rolling out the EU-ordered redesign of European search results, and a company official told Reuters it marks the largest reduction in service quality in Search's 29-year history. The changes strip real-time prices and other rich features out of the hotel, airline and restaurant carousels, and instead put one specialised

AI built a zero-click WeChat worm in two days, Tencent says

The Guardrails

AI built a zero-click WeChat worm in two days, Tencent says

Two stories this morning say the same thing from opposite directions: AI is now fast enough to break things faster than institutions can make rules about them. Security researchers at Calif say they used AI to find a memory-corruption bug in WeChat's calling code and write a working remote-code-execution exploit in about two days — and that Tencent has now mitigated it. The result is WeWorm, described as the first zero-click worm that spreads through WeChat calls across both iOS and Android: th

Open Source Radar — September 8: the browser is the new agent runtime

The Stack

Open Source Radar — September 8: the browser is the new agent runtime

Today's board is unusually coherent — three of the four hottest repos are about getting an agent onto the open web without paying Chrome's memory tax or getting fingerprinted at the door. The fourth is a research agent that wants to do your reading for you. Lightpanda (Zig, 35,096 stars) — A headless browser written from scratch in Zig rather than forked from Chromium, aimed squarely at AI agents and automation. The pitch is measurable: its own benchmark across 933 real pages puts peak memory a

IFM audits its own benchmark scores — and finds them 3.4 points too high

The Frontier

IFM audits its own benchmark scores — and finds them 3.4 points too high

Two releases worth your attention this morning: a fully open six-model fleet from Abu Dhabi's Institute of Foundation Models that ships with an audit of its own benchmark numbers, and a robotics hire that says where embodied AI thinks the bottleneck now sits. IFM published its own reward-hacking audit alongside K2 Horizon, and the flag rate is the most interesting number in the release. MBZUAI's Institute of Foundation Models launched six Apache 2.0 models — 0.9B, 3.7B, 7B, 32B, a sparse MoVA 3

Deep Dive — 800M parameters just beat a 122B model at African translation

The Frontier

Deep Dive — 800M parameters just beat a 122B model at African translation

The most interesting model release of the week did not come from a frontier lab. Tether's AI research arm shipped three offline translation models covering 19 sub-Saharan African languages, and the smallest one — 800 million parameters — reports better translation quality than a 122-billion-parameter Qwen and a 27-billion-parameter TranslateGemma across three public benchmarks. It runs on a phone with no signal, no account, and no per-token bill. That result deserves the standard suspicion ever

TSMC and Samsung commit to ASML's $400 million High NA EUV machines

The Arena

TSMC and Samsung commit to ASML's $400 million High NA EUV machines

The chip supply chain just locked in its next decade. Two of the three biggest buyers of advanced lithography committed to ASML's High NA EUV tools on the same day, and AI is the reason both gave. The world's two largest chipmakers have both committed to ASML's High NA EUV lithography machines, ending years of speculation about whether the industry's most expensive tool would ever reach volume production. Samsung said it will introduce High NA into DRAM high-volume manufacturing in 2028 — an in

ByteDance is building a real-time spatial video model

The Frontier

ByteDance is building a real-time spatial video model

Two stories this morning that point the same direction: AI's next arena is the space you can walk into, not the text you read. ByteDance is developing a real-time spatial video model, and founder Zhang Yiming is personally running it. Bloomberg reported that the system builds on the company's Seedance video model and generates interactive virtual environments — worlds that change as the user moves and speaks — rather than pre-rendered clips. The target launch is as early as next month, though t

Perplexity's local agent runs on Alibaba's Qwen, not Claude or GPT

The Arena

Perplexity's local agent runs on Alibaba's Qwen, not Claude or GPT

Two stories this morning about where the work actually happens — and who owns the model underneath it. Perplexity's on-device agent, Portable Computer, runs on Alibaba's open-weight Qwen3.8-27B rather than Anthropic's Claude or OpenAI's GPT. The company built its own optimization layer — a post-trained "PPLX 27B" variant of Qwen, which it has published on Hugging Face — to handle local file processing, data analysis and coding on Nvidia's DGX Spark hardware. Perplexity, valued at more than $30

AI 101 — What is a world model?

The Stack

AI 101 — What is a world model?

Robotics companies shipped at least ten "world models" between May and August. The phrase is now everywhere in AI — and means three different things depending on who says it. A world model is a learned internal simulation of an environment: give it what you see and what you're about to do, and it predicts what the world looks like next. That's the one-sentence version. A language model predicts the next word; a world model predicts the next state — the next camera frame, the next joint angle, t

Unitree's world model just fought without a human holding the controller

The Frontier

Unitree's world model just fought without a human holding the controller

Two releases this morning, one regulatory deadline already in force — and all three point at the same shift: AI systems are being asked to hold up in the messy real world, not just in a demo. Unitree says its UnifoLM-X2-1.0 model drove a G1 humanoid through a fully autonomous sparring session — no teleoperation, no pre-baked move list. Until now, every robot fight you've seen had a person behind it: an operator's motion mapped to motors, or a script replayed on cue. Unitree claims this one runs

Arm wants an SAE-style capability ladder for robots

The Arena

Arm wants an SAE-style capability ladder for robots

Robotics is about to get its own version of "Level 4." Arm is betting that whoever defines the vocabulary ends up defining the market — and it brought 80-plus companies along. Arm has published a Robotics Capability Framework that sorts robots into six levels of sophistication, and it is explicitly pitching it as the SAE automation ladder for everything that isn't a car. Announced alongside an extension of Arm Total Design to physical AI, the framework grades a machine's capability against real

DeepSeek opens 150 engineering roles — and makes agent fluency mandatory

The Arena

DeepSeek opens 150 engineering roles — and makes agent fluency mandatory

Two signals from China this morning: the open-weight lab is hiring like a platform company, and a phone is about to ship with an agent as its primary interface. DeepSeek's Harness team is hiring roughly 150 engineers at once, and the job post makes fluency with AI agents a stated requirement rather than a nice-to-have. Cui Tianyi, who leads DeepSeek's Harness team, posted the call for senior backend and server-side engineers, saying the volume of data, machines, training jobs and evaluation job

See all