DeepSeek R1 Is Awesome And Here Is Why!

Hi everyone! It's been a while without uploading any new videos due to my personal reason. Today we’re diving into DeepSeek R1—this open-source LLM that’s shaking things up! 🤯 Let me tell you why it’s low-key genius: It’s built ENTIRELY on Reinforcement Learning (no boring SFT here!), learns by trial-and-error like a digital daredevil, AND it’s been distilled into a lean, mean AI machine. Think of it like… a genius crammed into a tiny hoodie. 🧠✨

BUT… (yeah, there’s a ‘but’) is it ChatGPT slayer? Eh… not yet. GPT-4 still flexes harder in some areas. 💪 Oh, and hot take: This tech’s rise might rattle US stocks? 🚨 ‘Cause America’s obsessed with finance bots, not making stuff. 📉 Let’s unpack ALLLLL of this chaos.

Don’t forget to smash like if you’re hyped for open-source AI—or roast me in the comments if I’m wrong. Let’s go! 🔥"
#deepseek #ai #genai
---------------------------------------------------------------------------------------------------------------------
⌛️Time stamp:
00:00 introduction
00:49 background information about DeepSeek's CEO!
02:12 why being open-source is important to llms in the long terms?
03:55 reasoning-oriented reinforment learning vs. supervised fine tuning
05:50 chain of thought
06:30 model distillation
07:08 my concern
09:06 ending
---------------------------------------------------------------------------------------------------------------------
🌎Social media:
Welcome to reach out!
instagram: www.instagram.com/monicaaaaalxy/profilecard/igsh=a…
linkedin: www.linkedin.com/in/monicalin725