AI diagnoses major cancer with near perfect accuracy | Charles Darwin University https://www.cdu.edu.au/news/ai-diagnoses-major-cancer-near-perfect-accuracy
“HELM has a new leaderboard: HELM Capabilities v1.0! We curated 5 challenging datasets (MMLU-Pro, GPQA, IFEval, WildBench, Omni-MATH) and evaluated 22 top language models: https://x.com/percyliang/status/1902890719985160471?s=46
“We saw clear leaps in improvements over the previous ChatGPT-4o release: – Math #14 -> #2 – Hard Prompts #7 -> #1 – Coding #5 -> #1 https://x.com/lmarena_ai/status/1905340077339034104
“So what is the plan for the upcoming end of scientific publishing as we know it? Floods of AI-assisted, and eventually AI implemented, articles that look good (& may actually be good!) are already starting. There is the opportunity for something better, if we decide to build it” / X https://x.com/emollick/status/1902884726420959419
AI prediction model is a major breakthrough in weather forecasting – Earth.com https://www.earth.com/news/ai-prediction-model-is-a-major-breakthrough-in-weather-forecasting/
AI-driven weather prediction breakthrough reported | Artificial intelligence (AI) | The Guardian https://www.theguardian.com/technology/2025/mar/20/ai-aardvark-weather-prediction-forecasting-artificial-intelligence
“People are complaining it’s slightly inaccurate. That’s fair but to me that seems mostly an issue of the LLM part of it I am instead highlighting the impressive compositioning, text, overall flow of the diagram, etc. which was generated without needing to be specified at all” / X https://x.com/iScienceLuvr/status/1905067166350884941
YC-backed ReactWise is applying AI to speed up drug manufacturing | TechCrunch https://techcrunch.com/2025/03/17/yc-backed-reactwise-is-applying-ai-to-speed-up-drug-manufacturing/
“🏥🔗 Medical Supply Chain AI LangChain-powered system revolutionizing medical supply chains in low-income countries through intelligent monitoring and fraud prevention. Discover the future of secure medical supplies 👉 https://x.com/LangChainAI/status/1904201544305725749
“Built on our state-of-the-art Gemma models, TxGemma can understand and predict the properties of small molecules, chemicals, proteins and more. This could help scientists identify promising targets faster, predict clinical trial outcomes, and reduce overall costs. https://x.com/GoogleDeepMind/status/1905274926665208043
OpenAI exec leaves to found materials science startup | TechCrunch https://techcrunch.com/2025/03/17/openai-exec-leaves-to-found-materials-science-startup/
“The opposite: research has been finding an increasing “burden of knowledge” – we are learning so much that it is harder to master a field, so young scientists can be at a disadvantage in both research & entrepreneurship. By way of illustration: Roche’s maps of cellular process https://x.com/emollick/status/1902877533965586757
“For the record I do not bet on this multiyear research fad. To my understanding, the main way to manipulate the inner workings of AI is representation control. It’s been useful for jailbreaking robustness, finetuning resistant unlearning, utility control, model honesty, etc. 🧵” / X https://x.com/DanHendrycks/status/1905283397301452858
“Brain signals and LLM embeddings converge for predicting every spoken or heard word. Beautiful research from @GoogleAI They compared human brain activity during real conversations with internal embeddings from a speech-to-text LLM. Measured electrode signals in speech and https://x.com/rohanpaul_ai/status/1903373048260284868
“do not sleep on YOLOE 👏 this is the first monolithic real-time “detect and segment anything” model 🤯 it has text prompting mode, visual prompting mode and segment/detect anything mode my vibe tests went very well, see below ⤵️ https://x.com/mervenoyann/status/1904544071374311477
“@stanislavfort i am beginning to wonder if there is a builders/collators divide going on here. i think a lot of people want some sort of neat, orderly information space in which to navigate, while others are completely comfortable pushing the boundaries of what can exist in the first place” / X https://x.com/nptacek/status/1905200894645239988
“HELM has a new leaderboard: HELM Capabilities v1.0! We curated 5 challenging datasets (MMLU-Pro, GPQA, IFEval, WildBench, Omni-MATH) and evaluated 22 top language models: https://x.com/percyliang/status/1902890719985160471
