BlueSpork

I tested 17 uncensored local LLMs (8B to 32B) for compliance, accuracy, and logic. 🤖🔍 Here are the full test results: youtu.be/Yxb-9OWvjkI #UncensoredLLM #UncensoredAI

@i_Grga @openclaw This fixed the problem for me on a Mac curl -fsSL openclaw.ai/install.sh | bash -s -- --no-gum

@openclaw Having problem installing, INFO Installing OpenClaw (latest) inappropriate ioctl for device ✓ OpenClaw npm package installed WARN npm install openclaw@latest failed; retrying openclaw@next :(

🦞 OpenClaw 2026.2.21 ♊ Gemini 3.1 🔒 Massive security hardening 🎙️ Discord streaming + voice channels 🧵 Thread-bound subagent sessions 📱 iOS/Watch polish + gateway stability 🧠 Prompt caching tweaks 100+ fixes shipped while Karpathy called us a noun github.com/openclaw/openc…

@i_Grga @openclaw Yup, same issue

I really enjoyed this awesome video. It was created using Seedream 4.0 and WAN 2.2 youtu.be/jeZf2xvxr2M?si…

🚀 New video: Run SongBloom in ComfyUI – Local Suno Alternative 🎶 Step-by-step setup → GitHub clone, requirements install, checkpoints, and workflow. On my RTX 3090, a 2-min song took ~4 minutes to generate! Watch here 👉 youtu.be/-O7ZKZ2LibQ #ComfyUI #AIaudio #SongBloom

Infinite Talk: lip-sync/V2V (ComfyUI workflow) video/audio input -> video (lip-sync) On my RTX 3090 generation takes about 33 seconds per one second of video. Workflow: github.com/bluespork/Infi… video tutorial (step by step): youtu.be/LR4lBimS7O4

Infinite Talk: lip-sync/I2V (ComfyUI Workflow) image/audio input -> video (lip-sync) RTX 3090 - generation takes about one minute per second of video basic workflow: github.com/bluespork/Infi… video tutorial (step by step): youtu.be/9QQUCi7Wn5Q

@yukimamax DeepSeek R2: The Delay No One Saw Coming youtu.be/PzlqRsuIo1w

予想通りの展開きたー！習近平、中央政府のゴリ押しで開発停滞。 ここからは、なんで成果が挙げられないんだ、ということで粛清され、中央から管理者が送られて開発者のモチベが終わり、瓦解するというパターンですね。 AI開発は米国一強へ…。

I asked 7 major LLMs to visualise their minds. Original prompt I used “Can you visualise your mind using animation and a single self-contained HTML file using WebGL with GLSL shaders”. And then I followed up with more prompts if the results sucked. youtu.be/DcWBnQcBkBg

Meta signed a partnership with Midjourney. Instead of acquiring them, Meta will license Midjourney’s technology, the secret sauce behind MJ’s unique aesthetic quality, and use it in future AI models and products. #AI #Midjourney

Deepseek R2 coming out ... when it gets more cowbell Liang Wenfeng has been sitting on R2's release because it still needs more cowbell Huawei chips couldn’t finish a run, so back to Nvidia (Ascend stuck on inference duty) youtu.be/PzlqRsuIo1w #DeepSeek #LocalLLM #AI

Everything You Need to Know about gpt-oss in Under 2 Minutes youtu.be/oLFcpIGmTkU #GPToss #GPTOSS120B #LocalLLM

I just vibe coded a whole iOS app in Swift (without having programmed in Swift before, though I learned some in the process) and now ~1 hour later it's actually running on my physical phone. It was so ez... I had my hand held through the entire process. Very cool.

Comparing LLM performance on RTX 3060 vs. RTX 3090 with 7B, 14B, 32B, and 70B models. See how VRAM limits and offloading affect speed and how much faster the 3090 is over the 3060. #RTX3090 #AI #LLM youtu.be/VGyKwi9Rfhk

Qwen QwQ 32B Q4 hardware requirements Full video youtu.be/ImUDHA4a1Bc

How did 7 well-known LLMs handle a nested animation coding challenge with collision detection? See Claude 3.5, Gemini 2.0 Pro, Grok 3 & more face the test! (And see who *failed*) Full video: youtu.be/hVXWP_toBfk #AI #CodingChallenge #LLM #ArtificialIntelligence #Tech

Ollama vs. LM Studio—what's better? ⚡ Ollama: Open-source, CLI, faster 🖥️ LM Studio: Closed-source, GUI, easier 📊 Ollama is 34.48% faster Full breakdown here: youtu.be/QGtkaDWJZlA #AI #LLM #Ollama #LMStudio

DeepSeek R1 Hardware Requirements Explained youtu.be/5RhPZgDoglE #DeepSeekR1 #DeepSeekAI

@mtgreenee I belive Republicans should replace Trump with another candidate at the RNC convention. @mtgreenee

If I see any account saying we need to replace Pres Trump with another candidate at the RNC convention, I will unfollow you. You ARE the problem. Go join the Democrats because we do not want you. We are replacing you with new voters that won’t put up with dictator style banana republic shit.

@mtgreenee I belive Republicans should replace Trump with another candidate at the RNC convention. @mtgreenee

I will vote for President Trump even if he’s in jail!! I have never been more MAGA than I am right now! #Trump2024