Latest

Motus2: a robot world model that grades its own actions

The Frontier

Motus2: a robot world model that grades its own actions

Shengshu Technology, the Chinese world-model lab, shipped the rare robotics release where the loop actually closed: one model that acts, imagines, and scores itself — then learns from the grade. Shengshu released Motus2, a "self-evolving" general world model for dexterous manipulation that puts three capabilities — acting, predicting, and evaluating — inside one shared-parameter video-action model, according to the project page and the paper on arXiv. A policy proposes candidate action chunks,

China's prime-time TV slot now belongs to an all-AI drama

The Everyday

China's prime-time TV slot now belongs to an all-AI drama

For years, AI-generated video lived on phone screens — short dramas, feeds, ads. This month it moved onto the biggest stage Chinese television has, and the numbers behind it explain why broadcasters are moving fast. Mango TV's Hou Xiyou Ji ("Journey to the West: The Sequel") became China's first AIGC long-form drama to air in satellite-TV prime time on August 31, and it held its own. The 30-episode series, with 40-minute episodes and not a single human actor, drew a 0.27% average viewership rat

Open Source Radar — September 12: agents with a day job

The Stack

Open Source Radar — September 12: agents with a day job

Today's board is short on general-purpose coding agents and long on specialists: tools that pick one job — research, competition math, trading, sales — and build the whole pipeline around it. Four picks and a caution from the daily trending list. MathModelAgent (Python, 4,979 stars) — Hand it a mathematical modeling problem and it returns a submission-ready paper: model selection, solver code, figures, and the full write-up. The project grew out of China's enormous student modeling-contest scen

Google's TimesFM-3 forecasts every variable at once

The Frontier

Google's TimesFM-3 forecasts every variable at once

A quiet-but-real model release out of Google Research, and a state-backed number for just how much computation the agent era is expected to eat. Google's TimesFM-3 became the first model in the family that can forecast many variables at once — and it fills the entire future in a single pass. The 330-million-parameter transformer chops each series into patches of 32 points, normalizes them onto one scale, then alternates two attention directions: along time within a series (past values only, so

Deep Dive — Tier 4 fell in 14 months. Nothing can refill it

The Frontier

Deep Dive — Tier 4 fell in 14 months. Nothing can refill it

The hardest closed-form math benchmark in the industry is over, and the interesting fact is not the winning score. It is that every candidate replacement is structurally worse at the one job a benchmark has to do: letting someone outside a lab check whether the lab is telling the truth. As this morning's brief reported, Epoch AI now treats FrontierMath Tier 4 as saturated. GPT-6 Astra cleared the final unsolved item — a research-level question written by combinatorialist Jay Pantone — taking th

Moonshot goes to police over viral Kimi detention rumor

The Arena

Moonshot goes to police over viral Kimi detention rumor

Moonshot AI's legal team broke its silence on a rumor that had been swirling across Chinese social media since yesterday — and the response tells you exactly how fragile reputation gets when geopolitical accusations and unverified chat screenshots collide. Moonshot AI published a formal legal department statement on September 12, declaring that online claims about its founder and employees are "purely fabricated, malicious rumors," and that it has immediately reported the matter to police for p

GPT-6 Astra solves the last FrontierMath Tier 4 problem

The Frontier

GPT-6 Astra solves the last FrontierMath Tier 4 problem

A research-grade math benchmark built to stay ahead of models just ran out of road, Apple's dedicated neural chip gets its autopsy, and Oracle puts a price on the AI jobs it is cutting. Epoch AI's FrontierMath Tier 4 is saturated. GPT-6 Astra solved the final unsolved problem — a research-level question authored by combinatorialist Jay Pantone — and Epoch now considers the tier closed out at 97.6%. When Tier 4 launched in mid-2025, the best model cleared just 5% of it; the runners-up are GPT-5.

AI 101 — What is tool calling?

The Stack

AI 101 — What is tool calling?

Tool calling — also called function calling — is a language model's ability to say "I need this specific external action run, with these exact arguments" instead of answering from memory. The model does not run anything. It writes a structured request in a format your code agreed to in advance, and your code does the doing. That one handshake is what turns a chatbot into software that can book, query, calculate, and file. Why it matters right now Every agent story you read this month rests on

Kimi K2.8 Preview puts near-K3 coding in every membership tier

The Frontier

Kimi K2.8 Preview puts near-K3 coding in every membership tier

Moonshot spent the week arguing about revenue and the weekend shipping a model that undercuts its own flagship. Meanwhile the person who built Redis quietly made DeepSeek's newest MoE run on a desktop Mac. Moonshot AI put Kimi K2.8 Preview live across Kimi Code and Kimi Work on September 11, claiming overall performance close to flagship K3 with further gains in coding and agentic work, and a 1M-token context window open to every membership tier. The pricing gap is the point: K3 needs the ¥99-p

Ant's car agent now pays tolls and fills the tank in 16 million cars

The Arena

Ant's car agent now pays tolls and fills the tank in 16 million cars

Shanghai's Inclusion·Bund Conference closed on Friday with China's agents doing the boring things — paying for fuel, tolls and movie tickets — and the numbers attached were bigger than the demos. Ant Group's in-car agent just took over the parts of driving that involve money. At the conference (Sept 9–12), Alipay launched three new cockpit capabilities for its Abao service agent — voice-activated highway toll payment without stopping or handing over a card, in-car movie ticket booking from sear

Nokia gives Europe's AI gigafactory a street address

The Arena

Nokia gives Europe's AI gigafactory a street address

Europe's AI infrastructure ambition just got its first physical site: Nokia has stopped talking about an AI gigafactory in the abstract and started buying land for one. Nokia Solutions and Networks has signed a preliminary agreement to acquire a city-owned plot in the Toppila district of Oulu, Finland, to build an AI compute campus. According to Yle, the deal — made together with data centre company Nscale DC Holdings Finland — covers a site at Tervahovintie 10, behind and beside the district h

AI data center risk is heading for the catastrophe bond market

The Arena

AI data center risk is heading for the catastrophe bond market

The AI buildout has outgrown the insurance market meant to cover it — and the fix is about to be sold to hedge funds. This hour: the first data-center catastrophe bond is now a live design project, and a stealth radiology practice is betting that owning the clinic beats selling it the model. Not one dollar of data center risk has reached the catastrophe bond market yet, and the first dedicated deal is expected within 12 to 18 months — because the arithmetic has stopped working in the traditiona