“Qwen3.5-Omni Technical Report” Qwen3.5-Omni is a fully omnimodal model that understands text, images, video, and audio, then responds with text or real-time speech. The architecture they use is a Thinker-Talker split, where the Thinker performs multimodal reasoning, while the
RESEARCH
-
OpenAI announces ChatGPT Images 2.0 with updated intelligence and knowledge
By
–
Real-World Intelligence ChatGPT Images 2.0 has an updated knowledge cutoff of December 2025 and intelligence that allows it to expertly handle tasks end-to-end, from copywriting to analysis to design composition.
-
OpenAI announces ChatGPT Images 2.0 with improved accuracy and text rendering
By
–
ChatGPT Images 2.0 is a step change in detailed instruction following, placing and relating objects accurately, and rendering dense text, with the ability to generate across aspect ratios. It’s also accurate across languages and uses its expanded visual and world knowledge to
-
RFK Jr. seeks FDA approval for peptides without safety data, focusing on BPC-157
By
–
LLMs are still not consistent judges of qualitative work, and small changes to how that work is presented affect outcomes. Better harnessing and methods (multiple judging runs with randomized orders, etc) would certainly help, but the jagged frontier is very much still real.
-
Two New Approaches to Predict RNA 3D Structure
By
–
Predicting RNA 3D structure is a grand challenge of biology. 2 new approaches published today to get at it https://
sciencedirect.com/science/articl
e/pii/S1097276526002054?viaihub
… @MolecularCell https://
nature.com/articles/s4225
6-026-01223-x
… @NatMachIntell -
AI Security Threats and Economic Vulnerability at Scale
By
–
The doom might come, but it will be because of all the security holes that our current software has that AI can find fast. And attack at scale. The existential stuff could spring out of that. What if our entire economy got attacked at once? That's what gets me fearful. But
-
Softmax Creator Regrets Terminology Choice, Suggests Softargmax
By
–
John Bridle himself, the man who coined the term "softmax", says that he should have called it "softargmax"
-
Weiyao joins team to expand multimodal AI research work
By
–
Welcome Weiyao! so excited to have you work on expanding our multi-modal work along with @rown , @piotrdollar
, @liliyu_lili and others! -
OneVL: One-Step Latent Reasoning and Planning with Vision-Language
By
–
OneVL One-Step Latent Reasoning and Planning with Vision-Language Explanation paper: https://
huggingface.co/papers/2604.18
486
… -
Free AI/ML Learning Roadmap: Math to Production
By
–
Free AI/ML learning roadmap:
Math → ML/DL → specialization → production. Stats/Linear Alg/Calc: Khan Academy, 3Blue1Brown Practice: Kaggle Learn Core ML: Coursera ML, scikit-learn DL: http://
DeepLearning.AI, http://
fast.ai Frameworks: PyTorch/TensorFlow/Keras