Left to Right Pane one: deepseek-v4.1-flash on 4x RTX Pro 6000s
Pane two: glm-5.3-flash on 2x DGX Sparks Pane three: qwen3.8-next-flash on 2x DGX Sparks I'm testing all 3 on various tasks, mainly managing my meta-work, contracts, billing, status etc.. Qwen is very capable.
Just review this video of mine. My problem is not that there are no topics to choose from, but that there are too many topics to choose from.
I should choose the one with the widest audience. This issue is still too safe, and the audience is narrow. I will improve next time. The good thing is that I tried WorkBuddy's Deepseek V4.1 Flash and it really didn't work.
That's why I always greet the AI 🤖 I thank it, "great GPT, you overdid it Gemini, great Deepseek"
Andrej karpathy could have charged $2,000 for this course.
He put it on YouTube. The full training stack. Tokenization. Neural network internals. Hallucinations. Tool use. Reinforcement learning. RLHF. DeepSeek. AlphaGo. 3 hours of the most comprehensive LLM education that
You have reached the end of the archive
All of deepseek