Shopify Deploys Slack AI Agent River to Promote Company-wide Open Apprenticeship
Shopify has launched an internal AI agent named River, integrated into the company's Slack, capable of reading code, running tests, writing code, submitting PRs, querying data warehouses, and tracking production.
In the past 30 days, 5,938 employees have collaborated with River across 4,450 different channels, submitting 1,870 PRs in the main repository last week, with about one-eighth of the merged PRs written by River and manually reviewed.
River operates only in public channels, rejecting direct messages, and enforces that all interactions are searchable and observable, promoting knowledge penetration and collective learning across the company; this design transforms AI from a private tool into a visible "Lehrwerkstatt" (teaching workshop) for all, significantly increasing PR merge rates and organizational learning speed.
Source: Public Information
ABAB AI Insight
Tobi Lutke apprenticed at a Siemens subsidiary in Germany, learning programming by observing senior programmers in the basement, serving coffee, and immersing himself for long periods. This experience directly shaped River's design, continuing the German "Lehrwerkstatt" concept of learning on the job.
On the capital path, Shopify concentrates internal engineering resources on River's development and the company's knowledge base, allowing the AI agent to continuously absorb team best practices and context through public interactions, forming a collective intelligence amplifier rather than serving individual engineers, significantly reducing knowledge silos and training costs.
Similar to how the overall level of human players improved after the introduction of computers in chess, and how early open-source communities achieved rapid iteration through public code reviews, Shopify is currently in the early stages of transitioning from traditional private tools to a public AI apprenticeship workshop, with River becoming the core vehicle for organizational innovation.
Structural judgment: Essentially a technological replacement. River's enforcement of public interactions replaces traditional private AI windows, allowing the entire company to learn simultaneously from top engineers and AI through visible apprenticeships. The mechanism lies in eliminating private information barriers, making AI a collective mentor, and driving rapid compounding growth of knowledge and judgment within the organization.
ABAB News · Cognitive Law
Public interaction is not a risk, but a lever to turn the entire company into an apprenticeship workshop.
Private AI enhances one person, while public AI amplifies the whole.
Organizational speed depends on the lowest bandwidth channel; the public apprenticeship system allows AI to turn slow channels into compounding engines.