X Watchlist: AI/Local Inference Leaders

Source: https://x.com/0xSero/status/2035064089345478658 Captured: 2026-03-21

Accounts

#	Handle	Name/Role	Why They Matter
1	@karpathy	Andrej Karpathy	Best AI thought leader. Nanochat = best intro to training LLMs.
2	@steipete	Peter Steinberger	OpenClaw creator. Peekaboo, summarize.sh, oracle. GitHub is a treasure.
3	@badlogicgames	Mario (badlogic)	Mario's Pi = best simplest open-source agentic loop. Setting new open source standards.
4	@TheAhmadOsman	Ahmad Osman	GPU king. Giveaways, dense educational content on self-hosting/home inference. Connected to open weight labs, does interviews.
5	@sudoingX	sudoingX	Up and comer pushing limits of single GPU inference.
6	@Ex0byt	Ex0byt	Will be fundamental in making local inference on massive models possible.
7	@alexinexxx	alexinexxx	GPU kernel programming. Hard worker, learning in public.
8	@gospaceport	gospaceport	Hardware/homelab expert. Teaches hardware economics. Most impressive homelabs.
9	@alexocheema	Alex Ocheema	Founder of Exolabs. Pioneering Apple hardware inference. Mac mini/Studio expert.
10	@nummanali	nummanali	Prolific CLI tool builder -- LLM subscription budgets, Claude Code with alt models.
11	@thdxr	Dax (Opencode)	Opencode team. Good writer. Anti-doomer content.
12	@juliarturc	Julia Turc	LLM compression science. Best source for learning compression techniques.
13	@Teknium	Teknium	Nous Research & Prime Intellect. Hard-working, principled open-source teams.
14	@victormustar	Victor Mustar	Head of Product at Hugging Face. Enables publishing/sharing.
15	@louszbd	louszbd	Head of community at ZAI. Top open-weight LLMs. Supercharged the movement.
16	@SkylerMiao7	Skyler Miao	Making frontier intelligence fit on $10K hardware via MiniMax.
17	@crystalsssup	crystalsssup	Building best open-weight model. Releasing research before next-gen model.
18	@0xSero	0xSero	The curator. Local inference community builder.