BitPetro

Recruiting Navigators for the 3B³ Lab to beta-test the framework, regimes and predictive models and stress-test the 3B³ Framework in real-time. Join the thinkers building the "Rosetta Stone" of the post-ETF era. bitpetro.app/3b3lab/ #3B3Lab #SignalsFromTheLab #TheRebellionsLedger

Let's kick of Cowtown @bitblockboom at the #FortWorth @ftwStockyards . #FortBitcoin #BBB meetup.com/fortbitcoin/ev…

Join us for Fort Bitcoin's #56th meetup @ Rogers Roundhouse Tue Mar 25, 6:30PM! Beers, burgers, patio chats on mining, energy, sats. No agenda—just real BTC talk. All welcome 🍻⚡ RSVP: meetup.com/fortbitcoin/ev… #FortBitcoin #BitcoinerEvents

MiniMax M2.5 went fully open source. If you're running OpenClaw, this changes the game. It's basically Claude Opus performance but 95% cheaper. It scores 80.2% on SWE-Bench Verified. OpenClaw's been great for persistent AI agents with memory, tools, and messaging integrations. The problem is running it 24/7 with frontier models burns money fast. M2.5 solves that. 80.2% on SWE-Bench Verified versus Opus 4.6 at 80.8%. That's a 0.6% gap for a model that costs $0.15 per task versus $3. The architecture is what makes this work. 230B total parameters, only 10B active during inference. MoE model that gives you frontier performance without frontier compute costs. It was trained on 200K+ real-world RL environments across 10+ programming languages. During training, the model developed what MiniMax calls an "architect mindset." Before writing code, it plans structure, features, and UI design. This wasn't programmed - it emerged from RL. Tool calling is actually better than Opus. BFCL scores: 76.8% versus Opus at 63.3%. In practice, this means cleaner function calls and better multi-step workflows. Speed matches Opus too. M2.5 completes SWE-Bench tasks in 22.8 minutes versus Opus at 22.9 minutes. For OpenClaw users, the cost difference compounds. $1/hour for continuous operation at 100 TPS. Run agents 24/7 without burning your API budget. MiniMax also released MaxClaw - cloud-hosted OpenClaw + M2.5 with zero setup. One-click deploy, runs 24/7, no servers to manage. Same OpenClaw framework, same capabilities, powered by M2.5, running in the cloud. MiniMax M2.5 is fully open source. I wrote a full guide on setting up MaxClaw for automated research workflows. Link to the article in the replies.

Why Soviet Math Students Were Strong People often ask: Why were Soviet and Eastern European students so strong in mathematics? Why do Eastern European students dominate math olympics? Part of the answer lies in the textbooks. For more than a century, generations of students studied from Kiselev’s Arithmetic — a remarkable book known for combining: • mathematical rigor • crystal-clear explanations • carefully structured problems The book stayed in print for 125+ years and sold over 100 million copies, becoming a cornerstone of math education in Russia, the USSR, and China. Yet there was no proper English translation. I have now translated this legendary book into English. If you want to see how mathematics used to be taught when clarity and logic mattered, take a look: valeman.gumroad.com/l/arithmetic?u…

@kyle_e_walker The threats are real too, vibe carefully...

I configured my OpenClaw bot to fetch the most recent drilling permits from the Texas Railroad Commission. It grabs the permits, parses the data, maps them over satellite imagery, and sends them over to me via Telegram. The opportunities for AI + GIS in oil and gas are massive.

@kyle_e_walker Wait till you see the math I'm working on.

so cool: loc.closertotruth.com/interactive

I was today years old the first time I saw the "shape" of thought in 10,000 dimensions. #HDC #Neuroscience

Starlink’s now in Venezuela! So we can mine Bitcoin anywhere there’s energy.

The future of computing is not being about perfectly precise but about being vaguely right ten thousand directions. #POW coming soon... #HDC #NueralCompute

Imagine stepping into your embeddings that power RAG. 📚 In this proof of concept, you’re surrounded by courses. Semantically similar ones float right next to each other. I've built an XR prototype with Rémi at @itpocket where you can stand inside your knowledge base on Apple Vision Pro. Ask a question and the AI lights up the relevant documents in space, so you literally see where your answer comes from. When you ask something specific like "I want to be an astronaut, what can I learn", only a few tight clusters glow. But ask "I'm bored, what could be something fun to study" and clusters across the space light up. #applevisionpro #artificialintelligence #xr #rag