efebic

WTF?! New open-source physics AI engine absolutely insane! 🤯 Genesis is a new physics engine that combines ultra-fast simulation with generative capabilities to create dynamic 4D worlds for robotics and physics. TL;DR: 🚀 430,000x faster than real-time physics simulation, processes 43M FPS on a single RTX 4090 🐍 Built in pure Python, 10-80x faster than existing GPU solutions like Isaac Gym 🌐 Cross-platform support: Linux, MacOS, Windows, with CPU, NVIDIA, AMD, and Apple Metal backends 🧪 Unified framework combining multiple physics solvers: Rigid body, MPM, SPH, FEM, PBD, Stable Fluid 🤖 Extensive robot support: arms, legged robots, drones, soft robots; supports MJCF, URDF, obj, glb files 🎨 Built-in photorealistic ray-tracing rendering ⚡ Takes only 26 seconds to train real-world transferrable robot locomotion policies 💻 Simple installation via pip: pip install genesis-world 🤝 Physics engine and simulation platform are fully open-sourced 🔜 ”.generate” method/generative framework coming soon.

Vibe coders debugging an app they built with Claude Code:

What it's like being married to a 160+ IQ person: Sipping coffee and checking on our plants in the garden (our normal routine)... Mike: "Oh, I figured out how plants grow." Proceeds to give the best explanation ever, something I'd never heard or thought about in my life, even as a plant lover. His brain is always working and discovering new things, not just about coding and software, but about everything around him. He calls it First Principles thinking. He is amazing at ignoring all assumptions and conventions.

even if mythos is the name (rumored), and even if mythos primarily implies lovecraft (questionable), why would a model having an association with lovecraft be bad? out of all the claudes, the one most obsessed with lovecraft is, by far, Opus 3

Geoffrey Hinton did an MNIST subliminal learning experiment in the 2015 paper 'Distilling the Knowledge in a Neural Network'! They trained a neural network on a dataset that completly lacked the number 3 and it exactly what the number 3 looked like anyway. What's even crazier is that you can train a neural network entirely on nosie data with ghost logits that dont correspond to any real numbers and have it learn to classify all the numbers!

You all know that UNIX based systems have a really solid built in system of users, groups and permissions, right? You don't have to reinvent all of this stuff from scratch for AI agents

I’ve personally had issues with it growing up as I was introduced to it back when I was 17. I was lucky enough to inherit $16,000 from my grandparents yet stupid enough to gamble it away. It happened the worst way possible for me because I started out winning 10x my money from $500 to $5000 playing roulette online, and I did it slowly. Then just as slowly, I lost it all. In frustration I deposited 25% of my inheritance to win it back. Then lost it all again. It was 4am and I had school in just a few hours as I was sitting there with my laptop in bed. I was emotional, felt shame, and desperate to rid myself of those feelings I wanted to win it back and justified it by saying I would stop if I just got to breakeven, so I deposited another 25%, but this time I made a promise to myself that this was going to be it, no matter how painful. I didn’t play it slow this time, I bet it all in one bet, and lost it all. I can’t even begin to describe how terrified, shameful and disgusted I was with myself. Every fibre of my being wanted me to take the last remaining half and make my money back, but that night I overcame those urges and in a defining moment of my life stopped before it got too late. I never told my parents for obvious reasons, but if I did I could’ve gotten my money back on because it was illegal for them to allow me to play in the first place, but I never realised at the time and even if I did I’m not sure I would’ve been able to withstand the pain from telling them, and that’s how gambling sends you down a dark path very quickly. I don’t know where I would’ve been today if I didn’t manage to stop myself at the last minute. For that reason I’m personally deeply against gambling and I find it unethical, but I understand why people without that experience can think otherwise. I thought I would share to help others, it’s never too late to stop.

Google Meet >>>>> Zoom

PewDiePie didn’t “train his own LLM.” He fine-tuned an existing open-source model on coding benchmarks. His model started at 8%, crawled to 16% after format fixes, and one run hit 19.6% that briefly passed GPT-4o on a single benchmark before he couldn’t consistently reproduce it. The tweet makes it sound like a YouTuber casually built a frontier lab in his bedroom. What actually happened is more interesting: a guy with a $41,000 home rig of 10 GPUs and 424GB of VRAM spent months failing, retraining, and iterating on dataset quality until he squeezed marginal gains out of a fine-tune. This is the part worth paying attention to. The entire arc from October 2025 to now tells you where AI tooling has actually landed. PewDiePie went from building his first PC to running Qwen 235B locally, vibe-coding a custom chat UI, orchestrating multi-agent voting systems, and now fine-tuning models on custom datasets. He did most of this through AI-assisted coding itself. The video is literally called “I wish I never did this project.” He’s documenting how painful and tedious the process was. That honesty is the signal. The hype accounts strip that away and replace it with “what the f*ck, YouTuber beats DeepSeek.” The real takeaway: fine-tuning on specific benchmarks with curated data can let anyone temporarily spike a score past models that cost hundreds of millions to train. That tells you everything about how narrow benchmark gaming has become, and nothing about general capability. PewDiePie knows this. The people quote-tweeting him with shock emojis do not.

Revolution [a complete circle or turn round a centralpoint, axis] [to roll back, to bring back]