The absolute audacity of this PR spin is actually sickening. 🤮 Let’s be real: Anthropic gets designated a "national security risk" and effectively "banned "from the US government 48 hours ago because they wouldn't cave on the exact same "unrestricted use" demands. Now,

Market Profile chart right in the terminal.

Qwen 3.5 small models are dropping soon! Get ready for some exciting new developments. I have a feeling these small models are going to be a huge hit in the OpenClaw community. They are perfectly built for this. #Qwen #AI #OpenClaw

The strike rate, the shot selection, the composure under pressure. Sanju Samson just put on a clinic. Absolutely clinical performance to secure that semifinal spot. Sanju, you dropped this: 👑. Give that man his crown! 🏏🔥 #SanjuSamson #BleedBlue #KingSanju

Nano-banana 2 is shit, man. I simply copy-pasted my tweet and asked it to generate the image. First, it put the South African players, then I said, put the Sri Lankan players. Then it got the score wrong. WTF NB, you had only one job, lol.

Grok doesn't hallucinate. It injects itself with delulu.

🚀 Introducing the Qwen 3.5 Small Model Series Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B ✨ More intelligence, less compute. These small models are built on the same Qwen3.5 foundation — native multimodal, improved architecture, scaled RL: • 0.8B / 2B → tiny, fast,

Qwen 3.5-9B model just dropped on HuggingFace. Gonna be a new "small model" king. Looking at the model card: Huge jumps in "Terminal-Bench" and tool-use. Perfect for autonomous trading bots. (hello OpenClaw) 1M Context: Throw entire repos or massive datasets at it. Speed:

About Qwen-3.5-9B Benchmarks from HF model card Despite being less than 1/10th the size of the 120B model, Qwen 3.5-9B scores higher in MMLU-Pro (82.5 vs 80.8) and GPQA Diamond (81.7 vs 80.1). It means: It has better "reasoning" and factual accuracy in hard sciences and

Can't wait for Unsloth AI's fine-tuning script for Qwen-3.5-4B. Edit: It's already there. This is fast.

Has anyone tried Qwen 3.5's 0.8B and 2B models on a Raspberry Pi? I can't wait to run them on my Raspberry Pi 4B 8GB matchbox computer. #QWen35 #raspberrypi

Dubai Airport at the moment . Wealthy expats are willing to pay as much as half a million dollars per person to escape . But all aircraft is grounded . North , there is Afghanistan. East, there is Pakistan, West, there is Iran. South, there is desert. #USAirForce

Retail traders really dream of "buying the dip" at zero. If the Nifty actually crashes 100%, fiat currency is just expensive toilet paper and the entire financial system has vaporized. I won't be analyzing order flow; I'll be securing my perimeter and digging the bunker. And

Medvedev, Rublev, and 40 other ATP staff—including doubles champs Heliövaara and Patten, are currently grounded in Dubai. It’s wild that despite all the money and fame, they’re stuck just like everyone else. But what about the others? Human life has to matter regardless of

Bloomberg Terminal is just a $25,000-a-year subscription to an exclusive chat app, and the data analytics are just thrown in as a free bonus.

This is such heartbreaking news for the open-source community. Looking at tweets from his various colleagues (who also 'stepped out') seems like Alibaba pushed him out. The running theory is that Qwen may pivot away from open-source releases toward consumer-facing integration