DeepSeek has been out for over 2 months now, and things have begun to settle down. We take this opportunity to contextualize the developments that have occurred in its wake, both within the AI industry and the world economy. As systems get more "agentic" and users are willing to spend increasing amounts of time waiting for their outputs, the value of supposed "reasoning" models continues to be peddled by AI system developers, but does the data really back these claims?

Check out our DeepSeek minisode for a snappier overview!

EPISODE RECORDED 2025.03.30

• (00:40) - DeepSeek R1 recap
• (02:46) - What makes it new?
• (08:53) - What is reasoning?
• (14:51) - Limitations of reasoning models (why we hate reasoning)
• (31:16) - Claims about R1 training on Open AI
• (37:30) - “Deep Research”
• (49:13) - Developments and drama in the AI industry
• (56:26) - Proposed economic value
• (01:14:20) - US government involvement
• (01:23:28) - OpenAI uses MCP
• (01:28:15) - Outro

• DeepSeek website
• DeepSeek paper
• DeepSeek docs - Models and Pricing
• DeepSeek repo - 3FS

Understanding DeepSeek/DeepResearch

• Explainers
  • Language Models & Co. article - The Illustrated DeepSeek-R1
  • Towards Data Science article - DeepSeek-V3 Explained 1: Multi-head Latent Attention
  • Jina.ai article - A Practical Guide to Implementing DeepSearch/DeepResearch
  • Han, Not Solo blogpost - The Differences between Deep Research, Deep Research, and Deep Research
• Analysis and Research
  • Preprint - Understanding R1-Zero-Like Training: A Critical Perspective
  • Blogpost - There May Not be Aha Moment in R1-Zero-like Training — A Pilot Study
  • Preprint - Large Language Monkeys: Scaling Inference Compute with Repeated Sampling
  • Preprint - Chain-of-Thought Reasoning In The Wild Is Not Always Faithful

Fallout coverage

• TechCrunch article - OpenAI calls DeepSeek 'state-controlled,' calls for bans on 'PRC-produced' models
• The Verge article - OpenAI has evidence that its models helped train China’s DeepSeek
• Interesting Engineer article - $6M myth: DeepSeek’s true AI cost is 216x higher at $1.3B, research reveals
• Ars Technica article - Microsoft now hosts AI model accused of copying OpenAI data
• The Signal article - Nvidia loses nearly $600 billion in DeepSeek crash
• Yahoo Finance article - The 'Magnificent 7' stocks are having their worst quarter in more than 2 years
• Reuters article - Microsoft pulls back from more data center leases in US and Europe, analysts say

US governance

• National Law Review article - Three States Ban DeepSeek Use on State Devices and Networks
• CNN article - US lawmakers want to ban DeepSeek from government devices
• House bill - No DeepSeek on Government Devices Act
• Senate bill - Decoupling America's Artificial Intelligence Capabilities from China Act of 2025

Leaderboards

• aider
• LiveBench
• LM Arena
• Konwinski Prize
• Preprint - SWE-Bench+: Enhanced Coding Benchmark for LLMs
• Cybernews article - OpenAI study proves LLMs still behind human engineers in over 1400 real-world tasks

Other References

• Anthropic report - The Anthropic Economic Index
• METR Report - Measuring AI Ability to Complete Long Tasks
• The Information article - OpenAI Discusses Building Its First Data Center for Storage
  • Deepmind report backing up this idea
• TechCrunch article - OpenAI adopts rival Anthropic's standard for connecting AI models to data
• Reuters article - OpenAI, Meta in talks with Reliance for AI partnerships, The Information reports
• 2024 AI Index report
• NDTV article - Ghibli-Style Images To Memes: White House Embraces Alt-Right Online Culture
• Elk post on DOGE and AI