HackerTube

Building pi in a World of Slop — Mario Zechner

AI Engineer3 HN recommenders · 2026–2026
579K views · talk · news · Uploaded 2026-04-16

Mario Zechner explains why he built pi, a minimal coding agent, out of frustration with existing agent harnesses. He also discusses how AI agents are straining open source projects and argues for slowing down.

HN commenters recommend it as a first-hand argument for building your own lean agent harness instead of relying on bloated, vendor-controlled ones.

LLMs & generative AISoftware practice, open source & careers

Watch on YouTube

3 comments from Hacker News

@currywurst2026 · in “Zot now supports Claude Opus 4.8”

In 18min , Mario Zechner , the creator of pi will echo more or less your exact concerns as to why he developed it [VIDEO] Enjoy !

RecommendsView on HN
@kzrdude2026 · in “Claude Code uses Bun written in Rust now”

Mario Zechner's talk "Building Pi in a World of Slop" is a good listen, on the topic of adults in a "world of slop". [VIDEO]

RecommendsView on HN
@sbysb2026 · in “Harness Engineering for Self-Improvement”

There is a really good video by the author of pi.dev (which I have used to build some of my own harnesses): [VIDEO] The basis of the argument is that the labs are constantly pushing updates to their system prompts that are used in claude code or codex, which are exceptionally bloated and change the sand beneath your feet with every update. By rolling your own harness, as long as you keep up with the latest advances and changes in the ecosystem, you capture a lot of the 'control' that LLM-based development feels like it strips from you. Obvious disclaimer that I use pi.dev when I am aiming for consistency, not absolute quality. Custom harnesses on pi are what I ship, claude code is still my fallback if I need to make sure a PR is the highest absolute quality

Strongly recommendsView on HN