Opinions
Dario Amodei: ‘The Adolescence of Technology’
In a new essay, Anthropic CEO Dario Amodei argues that humanity is entering a critical period as powerful AI systems…
Opinions
NBER Study: 90% of Firms Report No Measurable AI Productivity Gain
A National Bureau of Economic Research study surveying nearly 6,000 executives across the US, UK, Germany, and Australia found that…
Opinions
Anthropic Study Maps How Claude’s Values Shift Across Models and Languages
Anthropic researchers developed a method to measure and compare the values that Claude expresses across different model versions and languages.…
Opinions
Ed Zitron: The AI Industry’s Economics Don’t Add Up
Media critic Ed Zitron argues that the artificial intelligence industry is built on an unsustainable economic foundation. He contends that…
Engineering
The ACE Lens: What Actually Makes Good Training Data for LLM Agents
A new paper proposes a unified framework for generating high-quality training data for LLM agents, representing agentic data as a…
Engineering
CritICL: Using Small Models’ Mistakes to Guide Bigger Models at Inference Time
New research addresses how to improve language model reasoning during inference without repeated generation or external verification systems. The proposed…
Engineering
GameWAM Combines World Modeling and Agent Control for Video Games
GameWAM addresses a gap between game-playing agents, which typically lack explicit modeling of world dynamics, and game world models, which…
Engineering
PAWBench Tests Whether Video Generation Models Are True World Models
A new benchmark investigates whether current video generation models can function as authentic world models by accurately reproducing the distribution…
Engineering
UrbanGround Tests AI Agents’ Navigation in a Digital Twin of Hong Kong
New research investigates whether multimodal large language models can translate local visual perception into effective spatial navigation within complex urban…
Engineering
TTPO: Test-Time Policy Optimization for Label-Free Reasoning
Researchers propose Test-Time Policy Optimization (TTPO), a label-free training objective for adapting language models to math reasoning tasks at test…
Engineering
Training Agents to Evolve with Their Harness: The TaoLive Digital Avatar System
A technical report from Alibaba’s Taobao Live team describes Harness-Aware Training (HAT), a method for training compact models to adapt…
Engineering
PILOT: A Supervisor-Worker Harness for Live Agent Self-Improvement
This paper introduces PILOT, a supervisor-worker agent harness that performs self-improvement live during a run rather than only after it…