Сrypto Gentlemen

🚨 OpenAI charges $0.006/minute. Google charges $0.024. AWS charges $0.024. Someone just open sourced a tool that does it for $0. And it's faster than all of them. It's called Insanely Fast Whisper. And that's not hype. That's the benchmark. 150 minutes of audio. 98 seconds to transcribe. On your own machine. No API key. No cloud. No per-minute billing. Here's what the numbers look like: → Whisper Large v3 + Flash Attention 2: 150 min of audio in 98 seconds → Distil Whisper + Flash Attention 2: 150 min in 78 seconds → Standard Whisper without optimization: 31 minutes for the same job → That's a 19x speedup. Same model. Same accuracy. Just faster. Here's what it does: → One command to transcribe any audio file or URL → Speaker diarization — knows WHO said WHAT → Transcription AND translation to other languages → Runs on NVIDIA GPUs and Mac (Apple Silicon) → Flash Attention 2 for maximum speed → Clean JSON output with timestamps → Works with every Whisper model variant Here's the wildest part: Otter.ai charges $100/year. Rev charges $1.50/minute. Descript charges $24/month. Enterprise transcription contracts cost thousands. Podcasters, journalists, researchers, lawyers, content creators — anyone still paying for transcription is lighting money on fire. 8.8K GitHub stars. 633 forks. MIT License. 100% Open Source. (Link in the comments)

This is the chart the AI circular financing bandwagon doesn’t want anyone to see. Why? Because it shows financing to feed the biggest ponzi scheme in history is drying up, expansion is taking a big hit and all the fake promises and lies will crumble miserably

Read Larry Fink’s Letter to Investors. 1blk.co/40KnIFr

Hyperliquid's third party ecosystem is now generating ~$100M in run-rate annual revenue. This is a significant milestone for Hyperliquid and speaks to the compounding impact of all the talent and capital pouring into its builder codes and HIP-3 ecosystem.

Bessent will soon double task to cap both oil prices and US Treasury yields - but having your cake and eat it too when pressure against you is building up on all directions is objectively tough

🛢️There's a lot being said about oil prices right now, so I put this chart together to help explain the major crude benchmarks and why they're all behaving differently. ⚪Brent (white) — The world's "default" oil price. Most global trade is priced off this. When the news says "oil is at $108," they mean Brent. 🟡WTI (yellow) — The U.S. benchmark, based on crude delivered to Oklahoma. It's the lowest line on the chart because American oil doesn't need to transit the Strait of Hormuz. 🟢Murban (green) — Crude from Abu Dhabi, delivered at Fujairah port, which sits just outside the Strait. Even though it technically doesn't have to pass through the chokepoint, drone strikes have hit Fujairah and nearby ports, pushing insurance and shipping costs up. 🟣Oman (purple) — The key benchmark for heavier crude sold into Asia. Many refineries in China, Japan, and South Korea are built specifically to process this grade. It's the highest line on the chart because Asian buyers are competing fiercely for a shrinking pool of cargoes. 🔴Dubai (red) — Used to price most long-term Gulf→Asia export contracts. It tracks alongside Oman as a measure of how hard Asian markets are being squeezed. The story isn't any single price — it's the gap between them. In late February these five lines were within $6 of each other. Now the spread between WTI and Oman is over $50. Since the U.S.-Israeli strikes on Iran began Feb 28, the Strait of Hormuz has effectively been closed. Daily transits have fallen from a historical average of ~138 ships to fewer than 5. The IEA has called it the largest disruption to global energy supply in history. Iran's IRGC has warned that not "a litre of oil" will pass for U.S. allies, while selectively allowing some Iranian, Indian, and Pakistani tankers through. Saudi Arabia is rerouting oil to its Red Sea port at Yanbu, and the UAE is using a pipeline to Fujairah — but combined pipeline capacity is only 3.5–5.5 million barrels/day vs the 20 million that normally flows through the Strait. Meanwhile, the 400 million barrel emergency reserve release by IEA members covers roughly 4 days of global consumption. Japan's refiners get ~95% of their crude from the Gulf. China receives 45% of its oil via Hormuz. South Korea, India, Thailand, Pakistan, and Bangladesh are all severely exposed. The wider the spread between the Asian benchmarks and Western ones on this chart, the more you're seeing that pain in real time.

A mysterious whale returned after 7 months of inactivity and spent 111.62M $USDT to buy back 50,706 $ETH at an average price of $2,201. 1 year ago, this whale sold 28,683 $ETH at an average price of $3,892 for 111.62M $USDT. What a perfect buy-low-sell-high move! intel.arkm.com/explorer/addre… intel.arkm.com/explorer/addre…

5 minutes ago, @karpathy just dropped karpathy/jobs! he scraped every job in the US economy (342 occupations from BLS), scored each one's AI exposure 0-10 using an LLM, and visualized it as a treemap. if your whole job happens on a screen you're cooked. average score across all jobs is 5.3/10. software devs: 8-9. roofers: 0-1. medical transcriptionists: 10/10 💀 karpathy.ai/jobs

1 million context window: Now generally available for Claude Opus 4.6 and Claude Sonnet 4.6.

Striking image from the new Anthropic labor market impact report.

February so far in AI… in just 26 days 👇 Feb 26 – Nano Banana 2 new version, faster and sharper Feb 25 Perplexity Computer AI that can actually use a computer to get things done Feb 24 – Claude Cowrok new Claude release focused on better workflow and execution Feb 21 - Grok 4.20 Grok update with stronger reasoning and real time edge Feb 19 – Gemini 3.1 Pro better coding, stronger agents, longer context Feb 18 – Google Lyria 3 AI that turns text, pics, or video into 30-second music tracks with lyrics Feb 17 – Claude Sonnet 4.6 faster model for coding and long reasoning Feb 17 – Fujitsu AI Dev Platform automates full software development Feb 16 – Qwen 3.5 cheaper, stronger agent-focused model Feb 12 – Gemini 3 Deep Think built for science and heavy research Feb 12 – MiniMax M2.5 open model, strong at coding Feb 12 – Seedance 2.0 realistic 1080p text-to-video Feb 11 – GLM-5 huge open model, strong reasoning Feb 10 – RynnBrain AI for robotics and the physical world Feb 5 – Claude Opus 4.6 major upgrade for coding and tool use Feb 5 – GPT-5.3-Codex faster coding-focused GPT Feb 5 – Kling 3.0 4K AI video with audio

Tom Lee's loss on ETH with other people’s money is now bigger than the losses incurred by users in the FTX collapse

Erik Voorhees(@ErikVoorhees), an early #Bitcoin supporter and founder of ShapeShift, is buying back $ETH. One year ago, he sold 11,616 $ETH($33.94M) at $2,922. Recently, he spent 20.38M $USDC to buy back 9,911 $ETH at $2,057. intel.arkm.com/explorer/addre…

China is physical AI. Robots with nunchucks.

13F nonsense Look at all these sophisticated investors dumping $IBIT in Q4. Are they just dumb noobs to crypto with paper hands or perhaps they were unwinding short $MSTR long IBIT trades. And taking massive profits.

People who don’t organize into tribes get wiped out by people who do.

Throughout Bitcoin's life span we have seen two indicators continue to be the best global market bottom signals: The Mayer multiple (distance from 200 day moving average) and the 200 week moving average. Both of these are clearly in long term accumulation territory. $BTC