Ethan B. Holland

“Announcing Evo 2: The largest publicly available, AI model for biology to date, capable of understanding and designing genetic code across all three domains of life. https://x.com/arcinstitute/status/1892248139333091577

Biggest-ever AI biology model writes DNA on demand https://www.nature.com/articles/d41586-025-00531-3

“Introducing Glass 4.0 – New The newest version of our AI clinical decision support platform features: – Continuous chat – Advanced reasoning – 275x expanded medical literature coverage – Increased response speed Try 4.0 now at https://x.com/GlassHealthHQ/status/1892574802327523360

“Evo 2, a DNA foundation model trained on 9T DNA base pairs, with state-of-the-art performance across a wide variety of biologically relevant tasks:” / X https://x.com/gdb/status/1892393483341631589

“When we start hill climbing on meaningful bio benchmarks it will be epic for the field and humanity.” / X https://x.com/RichardSocher/status/1894447036923351104

“Stanford open-sourced ‘MedAgentBench’, a new benchmark for medical AI agents It challenges agents with 300 clinically relevant tasks across 10 categories, requiring interactions with FHIR-compliant environments https://x.com/adcock_brett/status/1893708533645906281

“MedAgentBench presents an unsaturated agent-oriented benchmark that current state-of-the-art LLMs exhibit some ability to succeed at. The best model (Claude 3.5 Sonnet v2) achieves a success rate of 69.67%. However, there is still substantial space for improvement. https://x.com/jyx_su/status/1891995750835396926

“Google launched an AI ‘co-scientist’ to accelerate scientific discoveries It’s a multi-agent research assistant (built on Gemini 2.0) that generates and validates new hypotheses across areas like medicine, genetics, and more https://x.com/adcock_brett/status/1893708266544193827

“Google AI Research published an AI co-scientist paper . This multi-agent system follows a “generate, debate, and evolve” approach to generate scientific. Key ideas: – a multi-agent system with an asynchronous task execution framework – scaling test-time compute for scientific https://x.com/TheTuringPost/status/1895075839970324663

“NExT-Mol 3D Diffusion Meets 1D Language Modeling for 3D Molecule Generation https://x.com/_akhaliq/status/1892438110480302474

“Arc Institute and Nvidia dropped the largest AI for biology—in another open-source win! Trained on 9T+ DNA blocks from 128K species (entire tree of life), Evo 2 achieved 90% accuracy in predicting cancer-related gene mutations It can even design genomes https://x.com/rowancheung/status/1892507651000480049

“Arc Institute and Nvidia dropped Evo 2, the largest publicly available AI model for biology to date It’s trained on 9T+ DNA blocks from 128K species and achieved a 90% accuracy in predicting cancer-related gene mutations Another open-source win https://x.com/adcock_brett/status/1893708370709741788

“NVIDIA introduced CorrDiff for Earth-2, delivering kilometer-scale weather forecasts at 10,000x higher energy efficiency than traditional high-resolution modeling. ☁️ Generative Model CorrDiff uses diffusion-based AI to upscale coarse 25 km grids to 2 km resolution. It
https://x.com/rohanpaul_ai/status/1894894037582389763

“Big-Math: A Large-Scale, High-Quality Math Dataset for Reinforcement Learning in Language Models “In this work, we present Big-Math, a dataset of over 250,000 high-quality math questions with verifiable answers, purposefully made for reinforcement learning (RL). To create https://x.com/iScienceLuvr/status/1894348004272083385

“Helix coordinates a 35-DoF action space at 200Hz Controlling everything from individual finger movements to end-effector trajectories, head gaze, and torso posture! https://x.com/adcock_brett/status/1892579000817521092

“Today Gumloop rolls out 𝘭𝘪𝘵𝘦𝘳𝘢𝘭 magic ✨ Our new AI Web Research node finds you the answers to any question by scouring the web. -Is this company SOC2 compliant? -What university did this person go to? -What talks did they give? Millions of new use cases unlocked 🔓 https://x.com/gumloop_ai/status/1892664640103923742

“SynthLabs presents: Big-Math: A Large-Scale, High-Quality Math Dataset for Reinforcement Learning in Language Models https://x.com/arankomatsuzaki/status/1894232624203534657

“Bengio et. al. presents: Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path? https://x.com/arankomatsuzaki/status/1893855770321539458

“Towards an AI co-scientist https://x.com/_akhaliq/status/1894950342875369681

“This is a pretty serious engineering undertaking. Please join us to help build the future of Internet browsing with AIs doing deep research and tasks for us! https://x.com/AravSrinivas/status/1894069472262058434

“Over the past couple weeks I have spoken to experts who were skeptical about the value of AI in transforming white collar analytical work who changed their mind when exposed to Deep Research. It isn’t fully there yet, but I think this thread is an indication of why this is so.” / X https://x.com/emollick/status/1894020502919782646

“When using a Deep Research tool for the first time, you need to review the output with a critical eye: follow the links to make sure things are really cited, read every line for hallucination, etc. You aren’t going to keep that attention to detail long, so get an idea right away” / X https://x.com/emollick/status/1894645796047364248

“I am hiring postdocs at #KAUST to develop an Artificial Scientist for the discovery of novel chemical materials to save the climate by capturing carbon dioxide. Join this project at the intersection of RL and Material Science: https://x.com/SchmidhuberAI/status/1894419218231128424

“🚢 Deep research is rolling out today to all paid users! It can do week long research-oriented tasks in 15 mins. I’ve used it to better understand muon colliders, the renewable energy market, and AI post training techniques—and to research/purchase a basketball hoop for my kids” / X https://x.com/kevinweil/status/1894468278078357857

“Access to training data is a major barrier to doing high quality scientific research on AI (huge shout out to @AiEleuther @allen_ai @llm360 and others who train models on released data), but even more than that I feel the lack of interest in such things by AI researchers is.” / X https://x.com/BlancheMinerva/status/1894756271489724772

“deep research out for chatgpt plus users! one of my favorite things we have ever shipped.” / X https://x.com/sama/status/1894527988378550392

“Deep Learning architectures usually aren’t trained to perform search at test time, leading to sample inefficiency + poor generalization. Latent Program Network (LPN) builds in test-time adaption by learning a latent space that can be searched. @ClementBonnet16 @MattVMacfarlane https://x.com/ndea/status/1895184760403828967

“Organizations and firms change slowly, systems have a lot of momentum So, honestly, does science – but there is some evidence already that if have a string of automated discoveries from AI (not proven that is possible on a large scale yet), we can take advantage of them quickly.” / X https://x.com/emollick/status/1893758054350340540

“Generate an llms.txt for any website in seconds Excited to launch the @firecrawl_dev AI extension in @raycastapp – Along with the generator, you can also deep research and extract web data from your desktop Check this out 👇 https://x.com/nickscamara_/status/1894773159536062822

When AI Thinks It Will Lose, It Sometimes Cheats, Study Finds | TIME https://time.com/7259395/ai-chess-cheating-palisade-research/

Google’s AI co-scientist is ‘test-time scaling’ on steroids. What that means for research | ZDNET https://www.zdnet.com/article/googles-ai-co-scientist-is-test-time-scaling-on-steroids-what-that-means-for-research/

Scientists took years to solve a problem that AI cracked in two days https://macaonews.org/news/around-the-world/ai-superbugs-research-gemini-google-imperial-college/

Accelerating scientific breakthroughs with an AI co-scientist https://research.google/blog/accelerating-scientific-breakthroughs-with-an-ai-co-scientist/

“Game of life – Interactive – Empty space spawns new live cells – Old live cells at increased risk of sudden death – Death promotes birth of new cell – Cells leave postmortem trace Built with @v0 Inspired by @vercel 404 https://x.com/aymericrabot/status/1890261940871766470

Vevo Therapeutics Open Sources Tahoe-100M, the World’s Largest Single-Cell Dataset, as the Inaugural Contribution to Arc Institute’s New Virtual Cell Atlas | Arc Institute https://arcinstitute.org/news/news/arc-vevo

How Orakl Oncology is using DINOv2 to accelerate cancer treatment discovery https://ai.meta.com/blog/orakl-oncology-dinov2-accelerating-cancer-treatment/

“Microsoft ON A ROLL: BioEmu is a large-scale deep learning model for efficient prediction of biomolecular equilibrium structure ensembles 🔥 “Can sample thousands of statistically independent structures from the protein structure ensemble per hour” 🤯 Open weights on Hugging https://x.com/reach_vb/status/1892656772759916860