← back

Hyperspace More Nodes Is All You Need: Nicolas Schlaepfer

2.4K views · Sep 04, 2024 · 5:48 min · Watch on YouTube ↗
Takeaway

A decentralized peer network plus a fine-tuned planning DAG model can give power users an editable agentic workflow over diverse open models.

Summary

  • Hyperspace is a decentralized AI network using community-contributed personal-computer resources rather than centralized GPUs.
  • Existing AIOS product (Windows/Mac, built on llama.cpp) routes inference across the peer network.
  • New Hyperspace product fine-tunes a planning model that emits agentic DAGs in JSON from a user query, then renders them in a React-Flow node editor for user editing.
  • Uses Qwen 2 Instruct as reasoner and Llama 3 70B as summarizer per node, with Puppeteer + Beautiful Soup for in-house web scraping to markdown.
  • Pitches diverse open-source models as the answer over a single closed monolithic model, exposing core agent primitives (memory, planning, code execution, file system).
agentsdecentralizedplanning
Original description
Hyperspace is launching a new kind of agentic planning model, alongwith a breakthrough new AI application for power users. This is meant to run on the Hyperspace peer-to-peer AI network: the world’s largest and most advanced Uber-like network of open source models running on consumer devices (with over 15,000 nodes registered so far). It’s time to put a dent in the universe.

Recorded live in San Francisco at the AI Engineer World's Fair. See the full schedule of talks at https://www.ai.engineer/worldsfair/2024/schedule & join us at the AI Engineer World's Fair in 2025! Get your tickets today at https://ai.engineer/2025

About Nicholas
AI Software Engineer | Prompt Engineer | Passionate about AI & Team Collaboration