I rotate between Kling and Veo. Usually prompting in both and seeing which one does it better. I almost forgot Sora existed until the news that they closed it. Lol
GENERATIVE AI
-
Memory Systems and RAG Limitations in AI Models
By
–
If I had to guess it's less decay and more that memories have naive RAG-like implementations, so you're at the mercy of whatever happens to retrieve in the top k via embeddings. They don't process you in aggregate and over time (probably compute constraints) so they struggle to
-
LLMs Overfitting to RAG Context: A Systemic Training Bias
By
–
(I cycle through all LLMs over time and all of them seem to do this so it's not any particular implementation but something deeper, e.g. maybe during training, a lot of the information in the context window is relevant to the task, so the LLMs develop a bias to use what is given, then at test time overfit to anything that happens to RAG its way there via a memory feature (?))
-
The AI Scientist Published in Nature: Fully Automated Research
By
–
The AI Scientist: Towards Fully Automated AI Research, Now Published in Nature Nature: nature.com/articles/s41586-0… Blog: sakana.ai/ai-scientist-natur… When we first introduced The AI Scientist, we shared an ambitious vision of an agent powered by foundation models capable of executing the entire machine learning research lifecycle. From inventing ideas and writing code to executing experiments and drafting the manuscript, the system demonstrated that end-to-end automation of the scientific process is possible. Soon after, we shared a historic update: the improved AI Scientist-v2 produced the first fully AI-generated paper to pass a rigorous human peer-review process. Today, we are happy to announce that “The AI Scientist: Towards Fully Automated AI Research,” our paper describing all of this work, along with fresh new insights, has been published in @Nature! This Nature publication consolidates these milestones and details the underlying foundation model orchestration. It also introduces our Automated Reviewer, which matches human review judgments and actually exceeds standard inter-human agreement. Crucially, by using this reviewer to grade papers generated by different foundation models, we discovered a clear scaling law of science. As the underlying foundation models improve, the quality of the generated scientific papers increases correspondingly. This implies that as compute costs decrease and model capabilities continue to exponentially increase, future versions of The AI Scientist will be substantially more capable. Building upon our previous open-source releases (github.com/SakanaAI/AI-Scien…), this open-access Nature publication comprehensively details our system's architecture, outlines several new scaling results, and discusses the promise and challenges of AI-generated science. This substantial milestone is the result of a close and fruitful collaboration between researchers at Sakana AI, the University of British Columbia (UBC) and the Vector Institute, and the University of Oxford. Congrats to the team! @_chris_lu_ @cong_ml @RobertTLange @_yutaroyamada @shengranhu @j_foerst @hardmaru @jeffclune
→ View original post on X — @sakanaailabs, 2026-03-25 16:21 UTC
-
Apple’s Deep Integration with Google’s Gemini Model Revealed
By
–
Apple's deal with Google goes way deeper than anyone thought. Apple doesn't just get to fine-tune Gemini, they have full access to the model inside their own data centers. That means they can distill (and are doing so) Gemini's knowledge into smaller models purpose-built for
-
Google launches Lyria 3 Pro with 3-minute tracks
By
–
Google launched Lyria 3 Pro, and upgraded music model, capable of generating up to 3 minute tracks.
— 🚨 AI News | TestingCatalog (@testingcatalog) 25 mars 2026
“The model now understands the architecture of music. This makes it possible to prompt for intros, verses, choruses and bridges + generate songs with more complex transitions.” https://t.co/R5jK7y4hy6 pic.twitter.com/UE3z3p45DVGoogle launched Lyria 3 Pro, and upgraded music model, capable of generating up to 3 minute tracks. “The model now understands the architecture of music. This makes it possible to prompt for intros, verses, choruses and bridges + generate songs with more complex transitions.”
-
The Future Will Be Weirder Than Anyone Expects
By
–
The future is going to be weirder than anyone expects
— The Rundown AI (@TheRundownAI) 25 mars 2026
pic.twitter.com/bLo6117keLThe future is going to be weirder than anyone expects
-
Gemini Deep Think Revolutionizes Scientific Research with Advanced AI
By
–
Gemini Deep Think: Redefining the Future of Scientific Research https://
buff.ly/5K3sIxP
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
The Challenges of Personalization in Language Models
By
–
One common issue with personalization in all LLMs is how distracting memory seems to be for the models. A single question from 2 months ago about some topic can keep coming up as some kind of a deep interest of mine with undue mentions in perpetuity. Some kind of trying too hard. [Translated from EN to English]
-
Lyria 3 Pro Now Available on Multiple Google Platforms
By
–
If you're ready to get into the studio with Lyria 3 Pro, here's where to access the model: — @GeminiApp for Google AI Pro/Ultra subscribers — @GoogleAIStudio and the Gemini API — @producer_ai — Vertex AI — Google Vids for Workspace customers + Google AI Pro/Ultra subscribers [Translated from EN to English]