Blog

FEATURED POST

July 2, 2026/5 min read

Code review used to mean reading the diff carefully. Now it means skimming a diff an agent wrote and clicking approve. The bugs that get past skim-review are different — and the review surface has to change.

Matthew Eisner

Matthew Eisner

Head of Growth

May 21, 2026/4 min read

The Gigacontext Threshold

Context windows are scaling toward a billion tokens. The retrieval-vs-window architecture call is already being made.

Matthew Eisner

Matthew Eisner

May 12, 2026/6 min read

Most AI Agents Have No Memory

Most AI coding agents treat failure as a signal to retry. A new paper from CMU, the University of Washington, and Arm shows what changes when agents actually remember and distill their failures: 3.12× speedup on a standard GPU kernel benchmark, and on one operation (RMSNorm), a kernel 1.75× faster than expert-written CUDA.

Idan Beck

Idan Beck

April 14, 2026/4 min read

Build Now

Capital and labor are decoupling. The cost of building software has collapsed while VCs figure out what to finance. There\

Idan Beck

Idan Beck

January 12, 2026/12 min read

The 4 Phases of Vibe Coding

I built a terminal-based Counter-Strike clone with a coding agent. 49K lines in a week. These projects go through 4 distinct phases, and understanding them is the key to effective AI-assisted development.

Idan Beck

Idan Beck

January 12, 2026/8 min read

The Bootstrapping Loop

What would you build if you could write a million lines of working code per day? AI-accelerated coding is the latest turn of computing\\

Idan Beck

Idan Beck

January 6, 2026/10 min read

Why Business Velocity Will Be Measured in Tokens per Second

Code velocity is becoming business velocity. AI removes the headcount constraint on software speed. The bottleneck shifts from typing to intuition—and that changes everything.

Idan Beck

Idan Beck

Have a system AI could finally make possible?
Talk to us about connectors, hardware-facing software, migrations, or operational systems where the validation loop matters.
Talk to us

Zerg Blog

Posts on AI systems, connectors, hardware validation, and production engineering loops.