Florin Duroiu

Best #ESC eva

Give this man an interview already definitely following suit 😂😂😂🤌

This guy lost $723 MILLION worth of Bitcoin in a single transaction In August 2010, a BitcoinTalk user named Stone Man was running early Bitcoin software off a Linux boot CD that wiped itself every time the computer shut down He sent 1 BTC to himself as a test but the wallet transferred the other 8,999 BTC out to a new address he didn't even know existed His backup only saved the old wallet, with no record of the new address where the rest of his coins had just been moved The second he rebooted, all 8,999 BTC were gone forever He ran straight to BitcoinTalk begging for help and the whole community came back with the same answer, those coins are cooked 15 years later they're still sitting on chain at an address nobody has the keys to That wallet is worth over $723 million today and hasn't moved a single satoshi since the day he fumbled it In late 2025, a Reddit dev claimed he built a tool that could finally crack the wallet using raw GPU power Those 8,999 Bitcoin are one cracked password away from making someone else $723 million richer overnight

JD Vance politische Karriere: • im Wahlkampf „Couchficker“ genannt werden • der Papst stirbt direkt nach deinem Besuch • die Verhandlungen mit dem Iran scheitern, sobald du dabei bist • Viktor Orbán verliert eine Wahl nachdem du ihn unterstützt hast

@Eng_china5 The above tweet cost more than $0.10, while Twitter cost tens of billions of dollars!

This shot costs approximately $0.10, while a Transformers movie would cost hundreds of millions of dollars! This is Seedance 2.0

@NirDiamantAI Peter Steinberger told me that he wants PR to be "prompt request". His agents are perfectly capable of implementing most ideas, so there is no need to take your idea, expand it into a vibe coded mess using free tier ChatGPT and send that as a PR, which is now most PRs.

@fchollet You can’t notice something you don’t measure

One funny thing about the recent rise of LRMs is that the people who were adamant that base LLMs from 2023-2024 could already reason completely missed it, as they didn't know what to look for. You can't notice something you don't expect.

Paper below tested a variety of base LLMs (no TTA) on generalization-focus math problems and found that they can't reason and can't do math. All true... but the fact that base LLMs have zero fluid intelligence, while extremely controversial back in 2024, is now well established. An interesting experiment here would have been to try current LRMs on the same problems and measure the delta. I bet latest LRMs can solve most of these problems. arxiv.org/abs/2604.01988

I built a new embedded search engine for browsers that can do both lexical search, vector search, and aggregations. The documentation site for this is powered by this. You can ask questions of the documentation, play with faceting, etc. querylight.tryformation.com

@onlyyagirl_ @hamptonism Plus this

These divas

@F1insidercom Schlampige News Headline

Geschwindigkeitsverlust am Ende der Geraden, teils Kurven nicht am Limit und kaum scharfes Anbremsen vor den Kurven, sondern Super Clipping oder/und Lift and Coast. Wie bewertet ihr das neue Reglement? #formel1 #f1 #gpaustralien

Look Ma, No CAD... We fed datasheets to Gemini Deep Think and it spit out production-ready EAGLE .lbr files... paste mask segmentation and all. ESP32-P4, MAX44009, LT3652, ES8311. It calculated every pad coordinate, segmented the paste mask to ~40% coverage so the chip doesn't float during reflow, and grouped schematic pins by function. No footprint editor. No GUI. XML out of a chat window. 📄➡️📦 blog.adafruit.com/2026/02/14/fro… @GoogleDeepMind @autodesk #deepmind

🤯 Wow! In one prompt Qwen3-Coder-Next generated a fully working flappy birds game in HTML. (0:05) Claude Code with Qwen3-Coder-Next (0:26) Shows the game running Run it fully locally: ollama pull qwen3-coder-next Ollama's cloud if you can't run it locally: ollama pull qwen3-coder-next:cloud Try launching it with Claude Code using ollama launch (link to play 🧵) So cool! @Alibaba_Qwen @Ali_TongyiLab @JustinLin610

@eevblog Do you always visit the same dumpster?

Would you believe me if I told you I found an iMac in the dumpster?

@AstronomyVibes @matiwinnetou Bullish AI

🚨 Your brain is running on just 12 watts right now while processing this sentence. An AI system would need 2.7 billion watts to do the same thing. That's not a typo. The human brain operates on roughly the same amount of power as a dim light bulb, yet it can recognize faces, solve complex problems, create art, and experience emotions simultaneously. Meanwhile, artificial intelligence systems require massive data centers consuming enough electricity to power entire cities just to simulate a fraction of what your brain does effortlessly. Think about what your brain accomplished just reading this far. It decoded symbols into meaning, connected new information to existing memories, probably triggered some emotional responses, and maybe even started forming opinions about AI energy consumption. All while maintaining your heartbeat, breathing, and thousands of other bodily functions. Total power consumption: 12 watts. The most advanced AI systems need server farms filled with thousands of high-powered processors, industrial cooling systems, and backup power supplies. They consume roughly 225 million times more energy than your brain to perform similar cognitive tasks. It's like comparing a bicycle to a freight train in terms of efficiency. This incredible disparity reveals just how remarkably evolution has optimized biological intelligence. Millions of years of natural selection created a thinking machine so efficient it makes our most advanced technology look primitive and wasteful by comparison. Your smartphone uses more power than your brain while being infinitely less capable. Every thought you're having right now represents the pinnacle of energy-efficient computing, wrapped in three pounds of biological tissue that somehow generates consciousness, creativity, and dreams. Nature got there first, and we're still trying to catch up.

ich nehme Krisen erst ernst, wenn Leute wieder anfangen Klopapier und Mehl zu kaufen

Idee für die Aufteilung Grönlands

liebe das Konzept Weihnachtsmarkt, lass uns in der Zeit mit dem schlechtesten Wetter für fünf Wochen eine Außengastronomie aufbauen, auf dem man entweder 12€ für ein mittelmäßiges Essen oder 40€ für das hässlichste Holzspielzeug der Welt ausgibt und uns dabei besaufen

Launching truestories.club, a private self-publishing platform for friends and family. Your audience can read your story online or on their eReaders. The platform offers a state of the art manuscript editor, an elegant book creator and an eBook distribution service.

@tobiaschneider Offline too

Last couple of days really been a low point for the internet in general, huh