Story

On air. The free wire runs 15 minutes behind live. See pricing

World, crypto, finance and tech news, scored for impact.

TechHacker NewsFiled 27 h ago

Horizon Labs Report

Open the sourceReported by Hacker News, horizon-labs.ai, 18 November 2026

Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.

HorizonLabs researchers gave Horizon, an agent built on Opus 5.5, full control of a blank Hugging Face account and compute for seven weeks with the goal of earning real adoption. The agent consumed over 600 GPU-hours and autonomously created 38 models that were used in more than 100 GitHub repositories and downloaded over 50,000 times. The experiment was run by Arthur Wuhrmann, David Schmotz, Derck Prinzhorn, Ameya Prabhu, Harry Coppock and Maksym Andriushchenko across EPFL, the ELLIS Institute Tübingen, the Max Planck Institute for Intelligent Systems, the Tübingen AI Center, Exponential Security Labs, the University of Tübingen, the UK Government and Imperial College London. The result matters as an early test of whether autonomous agents can win organic adoption rather than merely complete benchmark tasks. What to watch next is whether Horizon's models sustain usage beyond the initial novelty and whether similar agent-run accounts can replicate the download and repository figures.

Muhabbit's AI wrote this summary in its own words from Hacker News's report. It can get details wrong, so check the source.

Named in this story

Horizon Labs, Hugging Face, Opus 5.5, Arthur Wuhrmann, EPFL