zach's tech blog

LLMs don’t make chips better. Good engineers do.

A new use-case for approximate computing.

VCs seem to be taking fewer risks. Why?

Part 2: Quilter, and building a compiler for the physical world

Part 1: Diode, and connecting language models to the physical world.

And should you consider investing if they do?

Especially if you want to be an electrical engineer.

Part 2: Silimate and the power of AI in chip design.

Part 1: Primitive Instruments, and hardware at the speed of software.

How do we stop this from happening again?

In which zach learns to hack.

Also: co-working in NYC, and Normal Computing is hiring!

Tech hubs and tech hubris.

The Count of Monte Carlo

Building chip companies could be so much easier.

Tiny chiplets, big dreams.

Obfuscated Queries, explained.

And why does Anduril have so many products?

Could AI revive an architecture from the 1950s?

If you raise $120M, at least try to write an honest press release.