AI: DeepSeek R1

DeepSeek R1 Cloned for $30?! PhD Student STUNNING Discovery - YouTube

technique: using LLM to train specific SLM (Small Language Model);
$30 = 10 NVIDIA H100 hours (could run on aws)

DeepSeek might not be as disruptive as claimed, firm reportedly has 50,000 Nvidia GPUs and spent $1.6 billion on buildouts | Tom's Hardware

The fabled $6 million was just a portion of the total training cost.

DeepSeek, China, OpenAI, NVIDIA, xAI, TSMC, Stargate, and AI Megaclusters | Lex Fridman Podcast #459 - YouTube

The Illustrated DeepSeek-R1 - by Jay Alammar

Deep-dive into DeepSeek (Practical AI #302) podcast

Daniel’s blog post on DeepSeek

DeepSeek R1 on Hugging Face

DeepSeek

Anthropic CEO Reveals New Details About DeepSeek R1 - YouTube

DeepSeek vs. Open AI - The State of AI w/ Emad Mostaque & Salim Ismail | EP #146 - Moonshots with Peter Diamandis | Podcast on Spotify