I think the expression is “pulling the ladder”! All labs trained their models by distilling (at the very least distilling the web) which allowed them to become the fastest growing businesses in the history of humanity and now that they have armies of lawyers and lobbyists, they
MACHINE LEARNING
-
LTX 2.3 Video Model Launches on Krea with LoRA Training
By
–
LTX 2.3 is now live on Krea.
— KREA AI (@krea_ai) 30 avril 2026
this video model generates for 1/10th the cost of other top models and can learn any style or character – just upload some references and train a lora
try it now 👇 pic.twitter.com/9vvqzLDBbHLTX 2.3 is now live on Krea. this video model generates for 1/10th the cost of other top models and can learn any style or character – just upload some references and train a lora try it now
-
Olmix Framework for Data Mixing Throughout LM Development
By
–
Our thanks to everyone who came out to hear @MayeeChen dive into her paper "Olmix: A Framework for Data Mixing Throughout LM Development." Replay ICYMI live: https://
youtube.com/watch?v=sFIPst
z0cTc
… -
OCR-Memory: Novel Approach to Long-Horizon Agent Memory Storage
By
–
// OCR-Memory // Well this is a unique approach to store memory for long-horizon agents. Most of the agent memory systems compress trajectories into text summaries and hope the model remembers what matters. But that's where the information loss hides. Long-horizon agents need
-
AI-Driven De Novo Protein Design and Nanomachines
By
–
No less, enabling the design de novo proteins, a brilliant review @nature "Over the next five to ten years, we anticipate the design of sophisticated protein nanomachines and materials with functionality ranging far beyond that generated during natural evolution for a wide range
-
GenAI Makes Isoleucine Dispensable Rewriting Alphabet of Life
By
–
Not something you'd see everyday—changing the alphabet of life. All of life organisms are are built from 20 amino acids. Now genAI is enabling life to be built with 19 amino acids, making isoleucine dispensable. @ScienceMagazine https://
science.org/doi/10.1126/sc
ience.aeb5171
… -
ML Model Races AI Against AI in F1 Simulator
By
–
This is so cool@brcullum built a machine learning model to race @dvassallo's #vibejam F1 sim and ran thousands of simulations to become the fastest
— @levelsio (@levelsio) 30 avril 2026
Now we could have AIs competing against AIs in F1 races (like that IRL competition already does?) https://t.co/XgK9ZzaNqxThis is so cool @brcullum built a machine learning model to race @dvassallo
's #vibejam F1 sim and ran thousands of simulations to become the fastest Now we could have AIs competing against AIs in F1 races (like that IRL competition already does?) -
Self-Driving Robot Labs Could Replace Biologists, Study Sparks Debate
By
–
Will self-driving ‘#Robot labs’ replace biologists? Paper sparks debate
by Ewen Callaway @Nature Learn more: https://
bit.ly/4aoy3fM #Robotics #Engineering #ArtificialIntelligence #Innovation #Technology -
Evaluating Agentic AI: Insights from Brookings Institution
By
–
How can we best evaluate #AgenticAI?
by Elham Tabassi Ramayya Krishnan @BrookingsInst Learn more: https://
bit.ly/42uHMMD #GenerativeAI #ArtificialIntelligence #MachineLearning #MI -
Learning path for deep learning and backpropagation
By
–
Ca dépend quel modèles ? 🙂 des modèles pour faire quoi ? Si tu parles de deep learning, Si tu veux partir de la base je te recommande de comprendre comment fonctionne un algorithme de backpropagation pour commencer. J'avais fais ca jadis :