One of the most striking points from today’s keynote was the concept of the "Jagged Frontier." We’ve all seen younger, AI-native workers use tools like Claude or Gemini with incredible speed. But as the keynote pointed out, speed doesn't equal expertise.
Without deep background knowledge, a user is just a conduit for whatever the AI says. They can’t see the "jagged edges"—those specific, high-stakes areas where AI confidently hallucinates or lacks the nuance of a seasoned pro. As we move forward, the value of a human won't be in the output, but in the judgment applied to that output.
We’ve spent years giving AI "product context" (your Jira tickets, your Confluence pages). But the new frontier is Human Context.
- Are you stressed?
- Are you working in a high-risk industry like automotive or aerospace?
- Are you experimenting or executing?
The keynote challenged us: How can you put what you do into context? In high-risk environments, you don't experiment with safety, but you must experiment with the "back of the plane" processes. If you aren't experimenting with AI in your workflows right now, you aren't actually working—you're just producing "work slop."
Here is the most sobering thought of the day: Our organizational structures date back to 1855. Our Agile Manifesto is over 20 years old. Yet, we are looking at a world where developers and knowledge workers are suddenly 100x more productive.
We know how to manage people. We have no clue how to manage AI agents. As AI begins to handle the "lower-level" tasks, the workforce will shift toward making "interesting choices" rather than doing rote work.
If you’re here on-site, don’t miss the Team '26-specific ACH certification. It’s a 15-minute, 10-question sprint that you can only take here. It’s a great way to prove you’re staying ahead of these shifts.
The goal was never just to increase output. It was to increase the quality of our work and our lives. As we integrate more context faster, the future looks uncertain—but as they said on stage today: "It should be good."