Home
Notes
Archive
About
Home
Notes
Archive
About
Subscribe
Sign in
Ayaz on Tech & AI
A few take-aways from GTC 2026 Keynote
What I think was significant in Jensen's keynote address at GTC 2026
Mar 23
•
Ayaz
Is recursive self improvement closer than we think?
How OpenAI's Codex 5.3 and MiniMax's M2.7 built their own successors.
Mar 19
•
Ayaz
The Repo Frontier
An alternative, hypothetical look at the Middle East conflict.
Mar 17
•
Ayaz
2
Second-Order Effects: How Hormuz Threatens AI Infrastructure
A look at the second and third-order effects of the closure of the Strait of Hormuz has on AI infrastructure buildout.
Mar 11
•
Ayaz
2
2
Some thoughts on DeepSeek’s DualPath improvement
Exploring bottlenecks in Agentic inference.
Feb 28
•
Ayaz
Most Popular
View all
What's inside the world's fastest AI processor: Cerebras Wafer Scale Engine
Jan 15
•
Ayaz
4
How I learned to write and speak well.
Jan 2
•
Ayaz
5
What's inside the Nvidia Vera Rubin NVL72 Rack
Jan 9
•
Ayaz
2
Some thoughts on prefill and decode
Jan 5
•
Ayaz
4
What's inside Microsoft's Maia 200 AI Chip
Feb 3
•
Ayaz
2
Second-Order Effects: How Hormuz Threatens AI Infrastructure
Mar 11
•
Ayaz
2
2
Latest
Top
Discussions
Some thoughts on Interleaved and Preserved thinking modes in GLM
A path to better, more efficient multi-step agentic tool calling at the model layer.
Feb 19
•
Ayaz
Recursive self-improvement loops
Is continual learning for AI the final frontier?
Feb 11
•
Ayaz
1
Some thoughts on building a god
My notes from a podcast I saw that talked about what these AI companies are doing in private.
Feb 10
•
Ayaz
What's inside Microsoft's Maia 200 AI Chip
A look at the new AI Accelerator ASIC from Microsoft.
Feb 3
•
Ayaz
2
What's inside the world's fastest AI processor: Cerebras Wafer Scale Engine
A look at how Cerebras has achieved a fast AI training and inference chip.
Jan 15
•
Ayaz
4
What's inside the Nvidia Vera Rubin NVL72 Rack
A quick look at the Nvidia Vera Rubin platform for building AI factories
Jan 9
•
Ayaz
2
Some thoughts on prefill and decode
Why people in the inference space will be talking about them more.
Jan 5
•
Ayaz
4
See all
Ayaz on Tech & AI
Rambling on about Tech, AI, Culture, and everything in between.
Subscribe
Blog
My Blog
Ayaz on Tech & AI
Subscribe
About
Archive
Sitemap
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
