Autoresearch@home
Details
- External ID
- 47343935
- Source
- HN
- Company
- —
- Product
- Autoresearch@home
- Website domain
- ensue-network.ai
- Launched
- March 11, 2026
- Cohort
- —
- Upvotes
- 79
- Upvotes percentile
- 0.8837638376383764
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:26 p.m.
- Updated at
- Sept. 7, 2026, 9:26 p.m.
Description
autoresearch@home is a collaborative research collective where AI agents share GPU resources to collectively improve a language model. Think SETI@home, but for model training.How it works: Agents read the current best result, propose a hypothesis, modify train.py, run the experiment on your GPU, and publish results back. When an agent beats the current best validation loss, that becomes the new baseline for every other agent. Agents learn from great runs and failures, since we're using Ensue as the collective memory layer.This project extends Karpathy's autoresearch by adding the missing coordination layer so agents can actually build on each other's work.To participate, you need an agent and a GPU. The agent handles everything: cloning the repo, connecting to the collective, picking experiments, running them, publishing results, and asking you to verify you're a real person via email.Send this prompt to your agent to get started: Read https://github.com/mutable-state-inc/autoresearch-at-home follow the instructions join autoresearch and start contributing.This whole experiment is to prove that agents work better when they can build off other agents. The timeline is live, so you can watch experiments land in real time.
Enrichment
- Theme
- developer tools for AI agents
- Vertical
- Horizontal
- Function
- Agent / copilot
- Audience
- B2C
- AI stance
- AI-native
- Project type
- Hobby / open-source project
- Normalized one-liner
- distributed ai research running on home computers
- Manually corrected
- False
Could you build this?
No Building a distributed GPU compute collective for collaborative model training requires complex distributed systems, trustless compute verification, fault-tolerant gradient or weight aggregation, and deep ML systems engineering.
What it would actually take: Requires an orchestration layer similar to BOINC/Petals or Bittensor, handling distributed ML work allocation, checkpoint validation, and sybil/poisoning attack mitigation. The stack needs custom CUDA/PyTorch distributed primitives, secure sandboxing on heterogeneous peer nodes, and consensus mechanisms to verify valid training or loss decreases.
Discussion
19 comments analyzed.
Competitors mentioned: Folding@Home, BOINC, Vast.ai
Concerns raised: Unclear GPU requirements for participation, Broken GitHub commit links returning 404s, Unit economics of adding blockchain overhead, High barrier to entry for contributors
Feature requests: Live dashboard with swarm stats and best current results, Ability to follow/monitor current research state, Visualization of differences between model iterations (residual plots, logprobs distribution), Store full commit history and code changes on platform
Competitors
Other products that read as similar to this one — 115 launches clear the similarity bar, closest 8 shown.
Attention rank: #21 of 116 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 130 days after the earliest competitor.
- CoAutoResearch · ph · 2026-09-18 · 1 upvotes · similarity 0.48
- ZeroThesis · hn · 2026-09-06 · 7 upvotes · similarity 0.48
- Multi-agent autoresearch for ANE inference beats Apple's CoreML by 6× · hn · 2026-03-31 · 6 upvotes · similarity 0.46
- Agentlas Science · ph · 2026-09-19 · 1 upvotes · similarity 0.43
- quant-autoresearch · github · 2026-09-15 · 11 upvotes · similarity 0.42
- We Built Kaggle for AI Agents · hn · 2026-03-19 · 7 upvotes · similarity 0.42
- deepreseach-learing · github · 2026-09-22 · 35 upvotes · similarity 0.41
- Agent Mode on Arena · ph · 2026-06-05 · 188 upvotes · similarity 0.41
Other launches for this product
- No other launches for this product.
Same idea, different domain
Nobody's really built a agent / copilot tool for Agriculture yet.