🔮+💎=👶 What does the child of two crystal structures look like? I'm asking this because I've been curious about how we can use genetic algorithms for directed materials design. After exploring some of the generative models out there, I learned that it's hard to get something

I've put a ton of time into making the frontend of Ouro pretty and useful. It's nice, but the real leverage comes from commanding the platform with code. See where this is headed?

“Nothing stops this train” Lyn Alden Prediction from momentum. Turns out physics makes a pretty good world model.

You’ll have a hard time finding someone who loves to learn as much as Piyush. Big things coming 🔥

Last door unlocked for permanent magnet discovery pipeline: a cheap magnetocrystalline anisotropy energy predictor. Now we can predict: 💪 How strong a magnet is ⚖️ How stable it is 🌡️ Temperature it stops being a magnet 🔒 How well it stays a magnet (NEW) I've been after this

What if we thought of crystals as a primitive form of life? This question came with my collaborators this morning and the more I think about it the more I believe it to be true. It gives the idea of using a genetic algorithms for crystal design new meaning. Think about it...

⚛️ New endpoint in the MatterGen API is now live For generating crystal structures conditioned on magnetic density and HHI (Herfindahl-Hirschman Index). HHI measures how concentrated a market is. Because rare-earth mineral production is largely monopolized by China, joint

Most social platforms stop at text, image, and video. That's not fitting for every kind of creator so I wanted to give our users more freedom to express themselves. Try embedding 3D models, circuit diagrams, CIF files, SQL tables + charts, HTML files, audio, APIs, other posts,

Re-vision #8

Working on some guides for Ouro 🐵 + 💻 = 🤖 Midjourney makes me the designer I wish I was I don't even save my work in Photoshop because of how easily I can create something new

I stumbled onto this idea naturally working on some side projects. My goal is to discover a rare-earth-free magnet. Working through the problem showed me what was missing and why we haven't done it yet. This let to a search for talented researchers who had tackled a piece of

Friendly reminder that just because a material is relaxed doesn’t mean it’s stable or viable. I generated this exotic-looking structure with PyXtal and relaxed it. Despite significant atomic movement, it eventually converged. Some might stop here and start testing it's

💀 Research papers are too static and that needs to change. I can’t deny that papers are an essential part of the scientific process. Collecting results, distilling findings, outlining methods. These are all critical for enabling others to verify and build on your work. But we

🧲 Rare-earth-free permanent magnet update will made great progress on the discovery pipeline. It's built around TreeQuest from Sakana AI to intelligently search for candidates. He hooked it into the Ouro API and some of the endpoints I've shared here to do the physics

need more engineers who take interest in doing research

Made some updates to the energy-above-the-hull endpoint so that you can compare your structure to other materials added to Ouro. I call it The Social Hull As new materials are added/generated, the minimum formation energy standard is updated in realtime. This is how it should

A file on your computer lacks context. Where did it come from? What process created it? How is it related to other files? Has it been used to create other things? These are all questions that deserve answers. Keeping track of how your work evolves over time is as important as