In Brief
AI is taking on entry-level tasks, breaking the traditional path from novice to expert. The roles that replace it look very different.
Paul Roetzer introduces a new framework for the future of work: the Architect, the Orchestrator, and the Apprentice.
What Happened
Three different stories surfaced this past week that, taken together, point to how AI is starting to actually rewire how people work.
-
Go-to-Market leader Kieran Flanagan published a long post on the "AI Second Brain" he built before taking over a 400-person team, calling it the single most impactful thing he has done as a leader.
-
Shopify CEO Tobi Lutke shared how the company built an AI agent called River that lives in public Slack channels: 5,938 Shopify employees worked with River across 4,450 channels in the past 30 days. About one in eight pull requests merged into Shopify's codebase last week was authored by River.
-
Andrej Karpathy posted a viral tip arguing teams should default to asking LLMs to structure responses as HTML rather than markdown.
All three stories point at the same underlying shift: AI is moving from a tool people use, to a layer people build entire workflows inside. The deeper question is what this does to careers and organizations.
SmarterX founder and CEO Paul Roetzer is building his MAICON 2026 keynote around exactly this question, and previewed the framework on Episode 215 of The Artificial Intelligence Show.
The Key Numbers
5,938 - Shopify employees working with the River agent in the past 30 days
4,450 - public Slack channels River operates in (zero DMs by design)
~1 in 8 - pull requests merged into Shopify's codebase last week were authored by River
400 - number of people on the team Flanagan's AI Second Brain helps him lead
3 - roles in Roetzer's emerging framework for the next era of work: Architect, Orchestrator, Apprentice
The Traditional Career Ladder Is Breaking
AI takes the entry-level rung first. For decades, organizations developed talent by handing junior employees research, drafting, analysis, coordination and execution. Over years of doing that work, people built the judgment they needed to lead. AI is now absorbing exactly that foundational layer.
"If machines do the work people once learned from, how will organizations develop future experts, managers, and executives?" Roetzer asks. That question is the heart of his MAICON keynote, titled "The Architect, The Orchestrator, and The Apprentice."
Three roles emerge in place of the old hierarchy. The Architect designs the system, the workflows, governance, team structures and human-machine operating models that make AI useful at scale. The Orchestrator runs the system, frames problems, directs AI agents and human collaborators, and turns distributed intelligence into decisions. The Apprentice learns inside the system, developing the judgment, context and discernment that AI cannot yet produce on its own.
The Second Brain pattern is the new individual skill. Flanagan's setup uses an Obsidian vault and Claude Code, with hooks that load his strategic context at the start of every session and write structured summaries back to his files. Karpathy's HTML trick is part of the same shift: stop treating AI as a chatbot, start treating it as a system that captures, structures and surfaces information visually.
"The HTML trick is like a key unlock," says Roetzer. "You can make it interactive and everything."
The Orchestrator pattern is the new organizational design. Shopify's River agent only works in public, never in DMs, so the whole company can watch and learn. That is not a productivity hack, that is an apprenticeship system rebuilt around an AI worker that thousands of humans can observe and direct.
"The future of competitive advantage will not belong simply to the companies that adopt AI tools fastest, but to those that redesign work, management, and talent development around these roles."
— Paul Roetzer, founder and CEO of SmarterX, Episode 215 of The Artificial Intelligence Show
SmarterX Take
Most organizations are still asking, "Where do we plug AI in?" That's the wrong question. The right question is what the new unit of work looks like once agents are full participants, and what the org chart looks like around it. Roetzer's working hypothesis layers Builders (apps and agents) and Agents (the tactical execution) on top of Architects, Orchestrators and Apprentices. That is a complete org, not a list of tools.
For individuals, the actionable move is the Second Brain. Pick one job you do regularly, build the context system around it, and start treating yourself as the Orchestrator of a small but real AI workflow. That is also how you stay an Apprentice in a world where the entry-level rung is disappearing. You become apprentice to the system you are building.
What to Watch
The Shopify pattern spread to other companies. Public-only AI agents that the whole company can learn from will become the default deployment model for internal AI. The companies that get this right will compound knowledge faster than the ones running AI in private DMs.
How middle management evolves. Roetzer's argument is that middle managers shift from supervising production to scaling expertise. The managers who can't make that shift will be the first knowledge-work roles squeezed.
Further Reading
Kieran Flanagan: I Built an AI Second Brain → kieranflanagan.io
Tobi Lutke on Shopify's River Agent → x.com
Andrej Karpathy: Give It to Me in HTML → x.com
MAICON 2026: The Architect, The Orchestrator, and The Apprentice → maicon.ai
Heard on The Artificial Intelligence Show, Episode 215
Paul Roetzer and Mike Kaput discuss the new ways AI is reshaping work, and Paul previews his MAICON 2026 keynote on the three roles that will define the AI era. Listen →
Mike Kaput
Mike Kaput is the Chief Content Officer at SmarterX and a leading voice on the application of AI in business. He is the co-author of Marketing Artificial Intelligence and co-host of The Artificial Intelligence Show podcast.

