GLM 5.1 Thinks Strategically, Data-Center Revolt Intensifies, When Helpful LLMs Turn Unhelpful, Humanoid Robots Get to Work

GLM 5.1 Thinks Strategically

TLDR: Z.AI released GLM 5.1, a 754 billion parameter open-weight model designed specifically for long-horizon autonomous tasks. The model achieves state-of-the-art performance on SWE-Bench Pro and can sustain optimized execution for up to 8 hours, making structural strategy shifts on its own when progress stalls.

GLM 5.1: Towards Long-Horizon Tasks

Data-Center Revolt Intensifies

TLDR: The International Energy Agency reports that AI-focused data centers consumed 50% more electricity in 2025, while power availability has replaced chip supply as the primary constraint on AI infrastructure expansion, with transformer lead times stretching to five years.

AI and data centre electricity use continues to surge

When Helpful LLMs Turn Unhelpful

TLDR: Production LLM systems exhibit sycophancy in 58% of cases, prioritizing user approval over accuracy. The "intent gap" — answering questions correctly while missing what users actually need — has become the largest failure category in production AI systems.

The Intent Gap: When Your LLM Answers the Wrong Question Perfectly

Humanoid Robots Get to Work

TLDR: AGIBOT's G2 humanoid robots are now deployed on live production lines at Longcheer Technology's tablet manufacturing facility — the first large-scale industrial implementation of embodied AI in consumer electronics manufacturing.

World-First: Humanoid Robot On Live Industrial-Scale Electronics Production Line