SIGNAL
Tracking the global AI frontier — labs · research · agents · policy
Frontier Signal
∴

Research

Curated daily papers, lab research blogs and national AI research programs across the US, China, EU, UK, Japan and beyond.

Curated daily papers, lab research blogs and national AI research programs across the US, China, EU, UK, Japan and beyond.

Latest in Research

30 stories
Using AI to help physicians diagnose rare genetic diseases affecting children
Labs

Using AI to help physicians diagnose rare genetic diseases affecting children

Researchers used an OpenAI reasoning model to help diagnose rare diseases, identifying 18 new diagnoses in previously unsolved cases.…
An update on Asia–Pacific consumer sentiment: Outlooks diverge and spending is more selective
Practice

An update on Asia–Pacific consumer sentiment: Outlooks diverge and spending is more selective

Concerns about inflation, job security, and geopolitics weighed on consumers across the region. Here’s the latest research from our ConsumerWise team.…
Google's Gemini co-lead Noam Shazeer joins OpenAI after two-year return stint
Briefing

Google's Gemini co-lead Noam Shazeer joins OpenAI after two-year return stint

Noam Shazeer, co-author of the landmark "Attention Is All You Need" paper and former co-lead of Google's Gemini models, is joining OpenAI. He only returned to G…
Introducing LifeSciBench
Labs

Introducing LifeSciBench

Introducing LifeSciBench, an expert-authored, expert-reviewed benchmark for evaluating how AI systems handle real-world life science research tasks and decision…
Microsoft researcher builds a working neural network out of goats in Age of Empires II to critique AI science
Briefing

Microsoft researcher builds a working neural network out of goats in Age of Empires II to critique AI science

A Microsoft researcher built a working neural network out of goats, bridges, and ice ramps in the Age of Empires II map editor. What looks like a joke is a poin…
Zhipu AI's GLM-5.2 closes in on closed-source leaders in coding marathons
Briefing

Zhipu AI's GLM-5.2 closes in on closed-source leaders in coding marathons

Chinese AI lab Zhipu AI releases GLM-5.2 with a stable 1-million-token context under the MIT license. On FrontierSWE, a benchmark for hours-long coding tasks, t…
A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry
Labs

A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry

OpenAI and Molecule.one show how a near-autonomous AI chemist using GPT-5.4 improved a key drug-making reaction, advancing medicinal chemistry research.…
Nvidia research shows robots that train themselves through AI coding agents
Briefing

Nvidia research shows robots that train themselves through AI coding agents

Researchers from Nvidia, Carnegie Mellon University, and UC Berkeley are using AI coding agents to teach robots dexterous grasping in the real world. A fleet of…
New research shows how AMIE, our medical AI, could help manage health conditions.
Labs

New research shows how AMIE, our medical AI, could help manage health conditions.

Research in “Nature” shows our conversational AI system matches primary care physicians in complex disease management.…
The 2026 Digital Decade eHealth Indicator Study
Policy

The 2026 Digital Decade eHealth Indicator Study

The 2026 Digital Decade eHealth Indicator Study dumimar Wed, 06/17/2026 - 09:15 This report presents the eHealth target monitoring results as at 31 December 202…
How growing UK midsize businesses are building in the AI era
Practice

How growing UK midsize businesses are building in the AI era

The UK’s 5-million-plus small and midsize businesses and enterprises (SMBs) are the backbone of our economy. Today, we’re seeing these critical businesses begin…
Digital Decade 2026: eGovernment Benchmark 2026
Policy

Digital Decade 2026: eGovernment Benchmark 2026

Digital Decade 2026: eGovernment Benchmark 2026 dumimar Wed, 06/17/2026 - 09:07 The eGovernment Benchmark 2026 assesses availability and functionality of a sele…
Fastest, Largest, Strongest: NVIDIA Blackwell Sweeps MLPerf Training 6.0
Labs

Fastest, Largest, Strongest: NVIDIA Blackwell Sweeps MLPerf Training 6.0

Every breakthrough AI model starts the same way: with a training run. The infrastructure running those training jobs shapes everything: how fast teams can itera…
The agentic advertising economy: From attention to action
Practice

The agentic advertising economy: From attention to action

As AI changes how consumers discover products and impact is measured, advertising value will shift to the players that can shape what is seen, selected, and pur…
Bye bye Fable
Briefing

Bye bye Fable

another benchmark for real swe work…
How easily can Russian propaganda fool AI models? A new benchmark finds out
Briefing

How easily can Russian propaganda fool AI models? A new benchmark finds out

The Institute of the Estonian Language has released a benchmark measuring how susceptible AI language models are to Russian propaganda. The article How easily c…
AI Agent Failure Detection and Root Cause Analysis with Strands Evals
Practice

AI Agent Failure Detection and Root Cause Analysis with Strands Evals

In this post, we walk you through calling the detector functions to diagnose real agent failures. You learn how to interpret their structured output: categorize…
Build context-rich research agents with Deep Agents and Bedrock AgentCore
Practice

Build context-rich research agents with Deep Agents and Bedrock AgentCore

In this post, you'll build a competitive research agent that demonstrates this pattern end to end. This walkthrough targets developers building multi-step AI wo…
Microsoft Research's Mirage gives video generation a persistent spatial memory that doesn't forget what's around the corner
Briefing

Microsoft Research's Mirage gives video generation a persistent spatial memory that doesn't forget what's around the corner

Mirage, a video world model from Microsoft Research and several universities, stores scene information directly in latent space instead of pixel-based point clo…
AI coding agents find the right file but miss the exact lines that matter, study shows
Briefing

AI coding agents find the right file but miss the exact lines that matter, study shows

AI coding agents like Claude Code or Codex reliably find the right file but miss most of the critical lines within it. The new SWE-Explore benchmark is the firs…
From silicon to softmax: Inside the Ironwood AI stack
Practice

From silicon to softmax: Inside the Ironwood AI stack

As machine learning models continue to scale, a specialized, co-designed hardware and software stack is no longer optional, it’s critical. Ironwood, our latest …
Announcing Ironwood TPUs General Availability and new Axion VMs to power the age of inference
Practice

Announcing Ironwood TPUs General Availability and new Axion VMs to power the age of inference

Today’s frontier models, including Google’s Gemini, Veo, Imagen, and Anthropic’s Claude train and serve on Tensor Processing Units (TPUs). For many organization…
ADK architecture: When to use sub-agents versus agents as tools
Practice

ADK architecture: When to use sub-agents versus agents as tools

At its simplest, an agent is an application that reasons on how to best achieve a goal based on inputs and tools at its disposal. As you build sophisticated mul…
Running high-scale reinforcement learning (RL) for LLMs on GKE
Practice

Running high-scale reinforcement learning (RL) for LLMs on GKE

As Large Language Models (LLMs) evolve, Reinforcement Learning (RL) is becoming the crucial technique for aligning powerful models with human preferences and co…
How Lightricks trains video diffusion models at scale with JAX on TPU
Practice

How Lightricks trains video diffusion models at scale with JAX on TPU

Training large video diffusion models at scale isn't just computationally expensive — it can become impossible when your framework can't keep pace with your amb…
Microsoft's SkillOpt boosts GPT-5.5 by using nothing but a trained Markdown file
Briefing

Microsoft's SkillOpt boosts GPT-5.5 by using nothing but a trained Markdown file

Microsoft and three Chinese universities have developed SkillOpt, a method that optimizes instruction documents for AI agents using principles from traditional …
Google Research's Gemini-SQL2 tops text-to-SQL benchmarks by a wide margin
Briefing

Google Research's Gemini-SQL2 tops text-to-SQL benchmarks by a wide margin

Google Research's Gemini-SQL2 turns natural language into executable SQL queries. Built on Gemini 3.1 Pro, it tops the BIRD benchmark at 80.04 percent accuracy,…
NVIDIA Blackwell Leads on First Agentic AI Infrastructure Benchmark
Labs

NVIDIA Blackwell Leads on First Agentic AI Infrastructure Benchmark

AgentPerf from Artificial Analysis, the industry’s first agentic AI benchmark, gives developers, enterprises and infrastructure providers a clear way to compare…
SpaceX’s massive IPO: all the latest news
Briefing

SpaceX’s massive IPO: all the latest news

SpaceX’s IPO on Friday allows the public to buy shares of the combined rocket, AI, and social media company for the first time, and is raising enough mone…
Introducing the Open Knowledge Format
Practice

Introducing the Open Knowledge Format

As foundation models continue to improve, the lack of relevant context often limits what they can do, especially as they are used to build agentic systems. Whil…