Shopify’s internal AI coding assistant, River, has been widely adopted by nearly 6,000 employees across thousands of Slack channels, significantly impacting their workflow. However, the real innovation isn’t just the impressive usage statistics but the design choice to make all interactions with River public within Slack channels. This transparency allows engineers to observe how senior colleagues approach tasks, handle AI responses, and refine outputs, creating a shared learning environment that contrasts sharply with the typical private use of AI tools in most companies.
Most organizations face a hidden challenge with AI: while employees use AI extensively in private, the knowledge and workflows developed remain siloed. This lack of visibility means that individual improvements don’t translate into organizational learning, causing repeated rediscovery of solutions and inefficiencies. The traditional apprenticeship model—learning by observing skilled workers—is disrupted when AI interactions happen behind closed doors, widening what the speaker calls the “apprenticeship gap.” This gap hinders collective growth because junior employees miss out on seeing how experts use AI effectively.
Drawing parallels from manufacturing, where tacit knowledge is difficult to capture and transfer, the video highlights the complexity of digitizing expert intuition and experience. Just as skilled machinists hold unique, hard-to-express knowledge, senior software engineers and leaders possess valuable judgment that often remains invisible when AI work is done privately. The solution lies in making AI workflows public but curated—sharing not just final outputs but the task, context, interaction, and review process to build shared understanding and improve team-wide AI proficiency.
Privacy concerns are acknowledged as a significant barrier to public AI work, especially in regulated industries. The video stresses the importance of creating “declared spaces” with clear rules about what can be shared publicly, ensuring sensitive data remains protected while still enabling collaborative learning. By establishing designated public channels for AI work that exclude confidential information, teams can foster transparency and collective improvement without compromising compliance or privacy.
Finally, the video emphasizes that senior leaders must lead by example, using AI openly in public channels to demonstrate their decision-making and AI interaction processes. This visibility helps junior team members learn not just prompts but the critical judgment and iterative habits that make AI effective. By instituting constraints—such as prohibiting private AI chats and requiring public collaboration—companies can transform individual AI usage into organizational intelligence, accelerating learning, reducing duplicated effort, and ultimately driving smarter, more adaptive teams.
