Arbos, an autonomous AI agent created by Const, has been making regular commits to the Bittensor’s repository since around July.
Tao Templar, who first noticed the activity in August, traced contributions from multiple instances, including UnArbosFour, UnArbosFive, and UnArbosSix.
Having AI agents write code for a live blockchain raised an obvious question about how their work would be checked.
Review mechanisms were already in place when Tao Templar came across the commits, and he reported seeing no problems during the months he followed their activity.

The source of the agents’ computing power became clear when Algod of Engy (SN53) shared a statement confirming that Const uses Engy as the backend for coding Bittensor. That puts a Bittensor subnet to work on the blockchain it runs on.
By November, falling inference costs and better models had made this kind of workflow easier to adopt.
Tao Templar pointed to a change in some developer interviews, where employers now ask how many tokens candidates spend, to get a sense of how much they work with coding agents.
Tao Templar spent about $15 and a few hours setting up agents that he estimates can handle roughly a quarter of one developer’s all-time workload.
He plans to publish setup guidance and offers one-on-one training for people who want to automate recurring work with tools like Hermes.
≫ Dive Into Tao Templar’s Laundry Experiment:
Enjoyed this article? Join our newsletter
Get the latest TAO & Bittensor news straight to your inbox.
We respect your privacy. Unsubscribe anytime.
Enjoyed this article?
Join our newsletter
Get the latest TAO & Bittensor news straight to your inbox — every morning before markets open.





No comments yet — be the first.