Home
Notes
Archive
About
Home
Notes
Archive
About
Subscribe
Sign in
Ayaz on Tech & AI
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
January 2026
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
How do Language Processing Units (LPUs) achieve efficiency
My take on understanding the LPU architecture briefly.
Jan 4
•
Ayaz
1
1
How I learned to write and speak well.
You can change yourself if you truly want to.
Jan 2
•
Ayaz
5
December 2025
Thoughts on improving inference time compute
What it means for Nvidia to acquire Groq
Dec 25, 2025
•
Ayaz
1
1
Driving to the future while looking at the rear-view mirror
We need new constraints for how we live our lives in an AI world.
Dec 23, 2025
•
Ayaz
LLMs and flat feet
Now you can have a good second opinion on a medical diagnosis on your phone.
Dec 15, 2025
•
Ayaz
1
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
