Andrej Karpathy released autoresearch a few days ago, but I don't have any ML projects to try it on so I started to use it to tune some new AI skills with autoresearch all day in the background, I layered in the evals-skills from Hamel (https://t.co/pdQQsHi0oH) into the workflow
— George from 🕹prodmgmt.world (@nurijanian) Mar 18, 2026
i really like the way @sainingxie articulates his thoughts and builds elegant analogies around them. i’d strongly recommend watching the section of the podcast where he talks about research taste. one idea that stood out to me was his view that good research taste means being https://t.co/GgTMYNUOLD
— λux (@novasarc01) Mar 17, 2026
https://t.co/sdI6uMfLGs
— marv1nnnnn (@marv1nnnnn1) Mar 18, 2026
Been saying this for a while. SF is incredible for building, but the consensus problem is real. The best founders I meet are the ones closest to their customers, not closest to other startups. Wrote more on this below https://t.co/sdBU1yxtMl https://t.co/yV2tFKTgeK
— Michael Bloch (@michaelxbloch) Mar 17, 2026
big fan of langchain's technical comms. they have consistently taught me something I didn't know I needed to understand the pattern in OpenSWE is much broader than an internal SWE agent btw adjacent usecases will look more like “knowledge work in a sandbox” than software https://t.co/LUtctasH1E https://t.co/hvpiEmrrO9
— angela (@probablyangg) Mar 18, 2026
This is an awesome write up on harness engineering from @rohit4verse AI is not wrong, you just have not built the harness correctly https://t.co/rXvljNjEBn
— Aparna Dhinakaran (@aparnadhinak) Mar 18, 2026
Fantastic read on the current state and possible future of Agent Harnesses. https://t.co/Caplp7Mv81
— Ajay Kulkarni (@acoustik) Mar 17, 2026
