Anthropic's Opus 5 surprise
PLUS: Automate tasks with Claude’s ‘Record a Skill’ feature
Intel · Flux
Flux Intel
Ce dont le monde de l'IA parle en ce moment, tiré directement des sources — du plus récent au plus ancien.
2 306 éléments collectés · Page 38 sur 39
DARPA's big bet on a light-based quantum computer
PLUS: Neuralink patients drive wheelchairs by thought
Black Forest Labs teaches video AI to run robots
PLUS: Edit videos faster with AI (for non-editors)
Travis Kalanick's $1.7B computer for the physical world
PLUS: Rural New York school gets robot teacher
Caught cheating
models, writers, and routers
OpenAI’s cyber test escapes the lab
PLUS: Build your own AI SEO specialist with Gumloop
Lightricks/LTX-2.5
3 503 pointsOpen models recap: more on Kimi K3, Qwen 3.8, Xi's WAIC speech, distillation, the open-closed gap, and what's next
A podcast with Florian Brand.
par Nathan LambertGoogle’s Gemini lineup has a Pro-sized hole
PLUS: Sell a high-value AI workflow audit as a consultant
China's bold plan to knock an asteroid off course
PLUS: Anduril and Archer bring the ‘Thunder’
Better design than Fable
unsolved maths problems vs ai models, who'll win?
Claude disproves an 87-year-old math problem
PLUS: Make AI note-taking easier with a “Captain’s Log” setup
Kimi K3: The open-weights escalation
The global implications on the AI ecosystem.
par Nathan LambertJitRoy2024/Qwen_Img_Space
86 pointsControlling Reasoning Effort in LLMs
How LLMs Learn Low-, Medium-, and High-Effort Reasoning Modes
par Sebastian Raschka, PhDI've got an Inkling
links to read over the weekend
How to use GPT-5.6
new desktop app and hosted sites in ChatGPT
6 months to live for open models
The most serious test to date of open source AI’s viability is happening right now.
par Nathan Lambertdream2589632147/Dream-wan2-2-fp8da-aoti-preview-2
143 pointsGrok x Cursor
More Fable, GPT-5.6 and new ChatGPT voice
My thoughts on Fable
what I want from a harness
Rchoks/wan555
148 pointsmrfakename/minimax-h3-ultra-fast
229 pointsLatest open artifacts (#22): Zyphra, Cohere, and Poolside are expanding the breadth of the ecosystem
An assessment of the open ecosystem and the motivations behind releasing models
par Florian BrandUsing Local Coding Agents
Using Open-Weight Models in Local Coding Harnesses as an Alternative to Claude Code and Codex Subscriptions
par Sebastian Raschka, PhDGLM-5.2 is the step change for open agents
A capability threshold I've been carefully monitoring.
par Nathan LambertBanning Open Source AI Would Be A Mistake
This post was originally an op-ed co-authored with Kevin Xu of Interconnected for a general, non-technical audience.
par Nathan Lambertkrea/Krea-2-Turbo
1 161 pointspnemrow/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental-neo
186 pointsState of the blog, mid-2026
About 3 years since I started writing weekly.
par Nathan LambertFrontier post-training recipe review with Finbarr Timbers
"Interview" #18
par Nathan LambertWelcome to the AGI era of AI governance
It's a one-way door and we weren't ready for it.
par Nathan Lambertmoonshotai/Kimi-K3
11 275 pointsraubatz/fk9b-headswap
60 pointsClaude Fable 5 and new AI safety fables
One step further into the power politics of frontier AI systems.
par Nathan LambertLLM Research Papers: The 2026 List (January to May)
A curated roundup of notable LLM research papers that came out this year
par Sebastian Raschka, PhDFarewell Ai2
This was my last week at the Allen Institute for AI (Ai2), where I got the great privilege to work on the Olmo models, to grow, to learn, and to have broad lasting impacts.
par Nathan LambertOpen and closed models are on different exponentials
Where marginally higher intelligence drives value, and where it doesn't.
par Nathan LambertSome ideas for what comes next, May 2026
Gemini Flash 3.5, Mythos, open-closed balance, America's open-source surge, emerging power struggles and more.
par Nathan Lambertkulkas2pintu/wan555
2 016 pointsRecent Developments in LLM Architectures: KV Sharing, mHC, and Compressed Attention
From Gemma 4 to DeepSeek V4, How New Open-Weight LLMs Are Reducing Long-Context Costs
par Sebastian Raschka, PhDAimeeBingmouQu/ProtectBirds
1 042 pointsOnise/Qwen-Image-Edit-2509-LoRAs-Fast2
315 pointsfroggeric/Qwen-Fixed-Chat-Templates
1 655 pointsM3st3rJ4k3l/FLUX.2-Klein-Multi-LoRA
549 pointsMy Workflow for Understanding LLM Architectures
A learning-oriented workflow for understanding new open-weight model releases
par Sebastian Raschka, PhDComponents of A Coding Agent
How coding agents use tools, memory, and repo context to make LLMs work better in practice
par Sebastian Raschka, PhDk2-fsa/OmniVoice
1 259 pointsA Visual Guide to Attention Variants in Modern LLMs
From MHA and GQA to MLA, sparse attention, and hybrid architectures
par Sebastian Raschka, PhDshootstuff/flux-img2img-uncensored
215 pointsaet256/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental
153 pointsA Dream of Spring for Open-Weight LLMs: 10 Architectures from Jan-Feb 2026
A Round Up And Comparison of 10 Open-Weight LLM Releases in Spring 2026
par Sebastian Raschka, PhDFrameAI4687/Omni-Video-Factory
1 477 pointslinoyts/Flux2-Klein-Face-Swap
553 pointsCategories of Inference-Time Scaling for Improved LLM Reasoning
And an Overview of Recent Inference-Scaling Papers
par Sebastian Raschka, PhDQwen/Qwen3-TTS
2 232 pointsThe State Of LLMs 2025: Progress, Problems, and Predictions
A 2025 review of large language models, from DeepSeek R1 and RLVR to inference-time scaling, benchmarks, architectures, and predictions for 2026.
par Sebastian Raschka, PhDLLM Research Papers: The 2025 List (July to December)
In June, I shared a bonus article with my curated and bookmarked research paper lists to the paid subscribers who make this Substack possible.
par Sebastian Raschka, PhDprithivMLmods/Qwen-Image-Edit-2511-LoRAs-Fast
2 723 pointsselfit-camera/Omni-Image-Editor
2 570 points