Introducing Dan Kagan-Kans
We just published the first post by our newest writer, Dan Kagan-Kans.
Proof1 independent outlet
Key points
- I first encountered his work in February, when he published “The left is missing out on AI” for Transformer.
- In July, Dan was the first reporter in years to get an interview with Hans Moravec, the namesake of Moravec’s Paradox — the observation that tasks that are easy for people are often hard for robots, and vice versa.
- Dan would like to get to know Understanding AI readers!
- Like Kai last year, Dan will be supported by a fellowship from the Tarbell Center for AI Journalism.
Sources (1)
- [1]Introducing Dan Kagan-KansUnderstanding AI (Timothy B. Lee) · Oct 5, 07:11 PM
We just published the first post by our newest writer, Dan Kagan-Kans.
I first encountered his work in February, when he published “The left is missing out on AI” for Transformer.
Extractive summary: sentences quoted from the sources.
Hype check
Independently evaluated1 independent look since
- Evaluation · Oct 6, 2026Test-Time Adaptation of Quantized ViTs via Single-Pass Quantizer-Aligned Recalibration
- Follow-up · Oct 6, 2026Exact-Solution Volume and Length Generalization in Transformers
- Follow-up · Oct 6, 2026Stability of Measure-to-Measure Transformers on Sub-Gaussian Data
- Follow-up · Oct 6, 2026RefRoute: Decoupling Conditioning Cost from References via Compact Residual Conditioning and Spatial Routing
- Follow-up · Oct 6, 2026Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures
- Follow-up · Oct 6, 2026Later Is Better: Token Reduction for ViTs Under Distribution Shift
Before this
- Oct 5, 2026LiquidAI/d1-omni-600M
- Oct 5, 2026MC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion Transformers
- Sep 30, 2026Cloudflare/clef-flash
- Sep 29, 2026microsoft/AesCode-32B
- Sep 29, 2026Language Models for Text Classification: From Bag-of-Words to Jev
- Aug 26, 2026vllm-project/vllm v0.28.0
