I do think that better compaction and teaching the models to re-learn context post compaction if they are unsure solves the need for really long context windows to an extent. That said, GPT 5.2 and the codex variants both support 400K context window which is a step up.
LLMS
-
Introduction of LingBot-World: A New AI World Model
By
–
4/
LingBot-World, a world model that achieves industry-leading performance in video quality, dynamic fidelity, long-term consistency, and interactivity – learn about it here: -
LingBot-VLA: A Vision-Language-Action Model for Robotics
By
–
3/
LingBot-VLA, a vision-language-action model designed to serve as a “universal brain” for real-world robotics – learn about it here: https://
afp.com/en/infos/robby
ant-open-sources-lingbot-vla-universal-brain-robots
… -
Jamba2: Open Source Model for February On-Device Applications
By
–
Model to try out in February: Jamba2. Fully open source under Apache 2.0, includes 3B (dense) model for on-device apps, intended for grounded QA workflows that don't call for the heavy “thinking token” overhead of reasoning models. Further details:
-
Jamba2: Open Source Model for On-Device QA Applications
By
–
Model to try out in February: Jamba2. Fully open source under Apache 2.0, includes 3B (dense) model for on-device apps, intended for grounded QA workflows that don't call for the heavy “thinking token” overhead of reasoning models. Further details:
-
Play-testing the Genie 3 AI model
By
–
My pleasure! Dropped a new video with a lot more genie 3 play testing:
-
Exploring Multi-Agent Ecology and Shared Knowledge Scratchpads
By
–
Absolute bars: “Sites like moltbook function as a giant, shared, read/write scratchpad for an ecology of AI agents – how might these agents begin to use this scratchpad to a) influence future ‘blank slate’ agents arriving at it the first time, and b) unlock large-scale
-
In-depth analysis and interview on Google DeepMind’s Genie 3 model
By
–
I tested Google’s world model Genie 3… Then DeepMind told me everything
— Bilawal Sidhu (@bilawalsidhu) 3 février 2026
00:00 – Intro & Authoring Workflow
00:27 – Genie 3 Playtesting & Demos
05:33 – Interview w/ Google DeepMind (Genie 3 co-lead @jparkerholder and Sr. PM Diego Rivas)
06:54 – Wildest emergent behaviors… pic.twitter.com/7Aafs67kaYI tested Google’s world model Genie 3… Then DeepMind told me everything 00:00 – Intro & Authoring Workflow
00:27 – Genie 3 Playtesting & Demos
05:33 – Interview w/ Google DeepMind (Genie 3 co-lead @jparkerholder and Sr. PM Diego Rivas)
06:54 – Wildest emergent behaviors -

AI Coding Rant as System Prompt
By
–
I turned Andrej Karpathy's viral AI coding rant into a system prompt. Paste it into http://
CLAUDE.md and your agent stops making the mistakes he called out. ———————————
SENIOR SOFTWARE ENGINEER
——————————— -
Codex Introduces New AI-Powered Planning Mode
By
–
most important update after the app – Codex now has a dedicated Plan mode! 🔥
— Vaibhav (VB) Srivastav (@reach_vb) 2 février 2026
It senses what questions to ask and searches for possible answers to them too – works well in both wide as well narrow domain problems
you can try it directly by /plan in the app as well as the CLI https://t.co/5uJ1WXHUv4 pic.twitter.com/dZ0hDzdRx5most important update after the app – Codex now has a dedicated Plan mode! It senses what questions to ask and searches for possible answers to them too – works well in both wide as well narrow domain problems you can try it directly by /plan in the app as well as the CLI
