This "does not feel right yet" is something that happens when I'm building the mental model of the thing I'm trying to understand
Today, this takes too much time compared to how fast the code gets generated I'm trying to have mental models on demand, here's an example from
DeepSeek 4.1 Flash is a great model for computer use
We tried it on our eval set and it performed at a similar level to Gemini 3.8, which previously had the best balance between cost and accuracy On our set of real-world business tasks, it performs with accuracy similar to that
Today's Top 5 — Sep 30
1. Abu Dhabi Commits Tens of Billions to Bypass Hormuz as Kuwait, Qatar and Bahrain Report $15.2 Billion in Lost Oil Revenue 2. August PCE Inflation Comes In at 3.4%, Below Forecasts, and Traders Cut Fed Hike Odds to 35% 3. DeepSeek Open-Sources
DeepSeek built its model for $6.5 million.
US labs spent hundreds of millions on theirs. We sat down with Kevin T. Carter, founder and CIO of emqqglobal, on why China is now favored to win the AI race.. Timestamps: 0:44 Where the real AI money is right now 3:12 Cheap tokens
You have reached the end of the archive
All of deepseek