GLM-5 Tech Report is here! This new model from @Zai_org & @Tsinghua_Uni uses a more efficient architecture (DSA) to cut costs and a novel training system that lets the AI learn from complex, long-term tasks. It achieves state-of-the-art results on open benchmarks and
INNOVATION
-

NVIDIA DreamZero Enables Robots Master Complex Tasks New Environments
By
–
What if robots could master complex tasks in totally new environments without needing thousands of repetitive demonstrations? NVIDIA presents DreamZero to do exactly that. The researchers built a World Action Model that learns physical dynamics by predicting future video frames
-
Productivity Growth Projections for 2025 Double Historical Averages
By
–
The productivity numbers for 2025 have not been published yet.
I see the large downward revision in employment, combined with strong GDP growth, and I infer that when the productivity numbers are published, they will be about 2.7% for 2025.
That is about double the average over -

Abacus AI Desktop: Unified AI Coding and Desktop Assistant Platform
By
–
Abacus AI Desktop brings AI coding and a desktop assistant into one place. From a CLI and coding agent to chat, voice assistance, and a VS Code extension, everything works together in a single setup.
-
China’s Tech Advances Challenge U.S. AI Leadership Position
By
–
China's tech shock threatens U.S. AI monopoly: 'just getting started' https://
share.google/nB2tBtEYXMfQ2O
1B9
… #Tech #Chinatech #ustech #AI #bigtech #ArtificialIntelligence @PawlowskiMario @chidambara09 @Ym78200 @CurieuxExplorer @efipm @bigfundu @sayedflah @Ronald_vanLoon @cyngn @belindabeibi -
AI Code Generation Impact on Software Engineers at Anthropic
By
–
If AI writes 100 per cent code at Anthropic, what will engineers do? https://
share.google/7xGLEz5xrW1fxD
Ghc
… #Anthropic #vibecoding #vibecode #coder #Developpement #developer -

Cerebras Partners with OpenAI at Stanford Student Hackathon
By
–
Over the weekend, the Cerebras team connected with 1,000+ student hackers at Stanford’s @hackwithtrees
! Some highlights:
• Met so many students building with the Cerebras API and OpenAI's new GPT-5.3-Codex-Spark model (powered by Cerebras).
• Heard from our partner, OpenAI’s -
Agent Wizard by PolyAI builds voice agents from websites
By
–
The tool is called Agent Wizard by PolyAI. And they just opened a waitlist for Agent Wizard. You give it a URL. It reads your entire site. FAQs, product catalog, store hours, contact info, policies. Everything. Then it builds a voice agent that can actually answer customer
-
LLM Agent Time Horizons: New Analysis on Model Capabilities
By
–
Ever since @METR_Evals
' fascinating work on LLM agent time horizons, I've wanted to see other attempts to draw conclusions from the same data. In a separate approach by Fengyuan and Jay, we too infer an exponential, but with shorter horizons for recent models (~2h vs METR's ~5h) -

Google adds Revise feature to Slides on NotebookLM
By
–
Google released the "Revise" feature for Slides on NotebookLM, allowing users to customise individual screens with a prompt.
