Introduction

It’s a pretty exciting time in tech-land. AI buzz is everywhere, and while we can’t believe all of the hype, I think it’s clear at this point that there are some genuine, game-changing improvements it has brought and will continue to bring here at Disco.

This document is intended to outline our approach to AI usage at Disco, and to codify some of the rules and expectations we agree to internally around that use.

As with all of our policies, it’s a document open to comments and proposed adjustments from everyone at the company — we certainly don’t expect it to stay static.

Tools

There is no set list of “approved” AI tools at Disco — as with anything, if you find something useful for your job, use it!

Some of the tools we recommend and pay for include:

Claude.ai (everyone has access to Claude chat);
Claude Code (required for engineers, but open to anyone in the business);
Gemini (through our Google Workspace subscription).

Ethical considerations

There are a number of thorny ethical considerations around AI:

The environmental impact of the resources used to train and operate AI;
Issues of ownership and copyright related to training data;
The broader societal impacts of AI and the potential impact on job security, social cohesion, and all out robot war.