For the last three decades, we’ve been told that “software is eating the world.” But as we cross the threshold of 2026, the world is starting to bite back.
We are currently witnessing a dual-narrative collision that is rewriting the rules of industry. In one corner, AI agents are predicted to replace traditional SaaS products, leading to a “SaaSpocalypse” that has already vaporized hundreds of billions in market value. In the other corner, foundation models for physical AI—trained on hundreds of thousands of hours of real-world data—are promising to automate the manual workforce.
But here is the irony: While software is becoming “thinner” and more ephemeral, the value is shifting back to the “thick” reality of atoms, warehouses, and the physical messiness of the supply chain.
The Orchestration vs. Ownership Trap
In a recent episode of The Arqiv, Brent Hagan, Chief Supply Chain Officer at Lob and former Amazon Robotics GM, sat down with Steve Ryan to discuss this exact collision. Hagan’s career has been a masterclass in the friction between digital commands and physical execution.
The tech industry is currently obsessed with “agents”—AI that can write code, manage calendars, and buy subscriptions. But an agent cannot un-spill a jug of protein powder on a fulfillment floor. It cannot fix a “spinning” robot that has lost its traction on a dusty barcode.
The lesson for 2026 is clear: Software is no longer the moat; the physical orchestration is. Lob’s survival strategy in the age of AI isn’t to build more complex software, but to act as a “thick” physical layer. By using APIs to trigger physical mail and logistics, they’ve built a system where the software is a “thin trigger” for a robust, patent-protected supply chain. They don’t own the fixed assets, but they own the orchestration—the ability to make third-party networks mimic the reliability of a first-party giant.
The Human Element: The Whiteboard vs. The Dashboard
As we lean into automation, there is a growing disconnect between what the tech industry builds and what operators actually need.
We’ve all seen the “Digital Twin” demos at CES—pristine, AI-powered simulations of warehouses. But as Hagan points out, those models often assume static, perfect conditions. The real world is inherently imperfect.
-
The Dashboard Fallacy: High-tech dashboards can track every KPI, but they often fail to explain why a tool disappears on the night shift or why a robot got “lost” because a pallet blocked its path.
-
The Whiteboard Reality: The most innovative solutions often come from “Gimba”—going to the floor, seeing the mess, and hashing it out on a giant Post-it note with the people closest to the problem.
“You can’t automate the decision-making that happens when you get humans in a room to solve a problem that isn’t in the data.”
The Unexplored Angle: The Resilience Paradox
There is a missing piece in the AI conversation that we must address: The Resilience Paradox. As we use AI to optimize supply chains for 99.9% efficiency, we are inadvertently making them more brittle. An AI agent can find the cheapest freight lane in milliseconds, but it lacks the “common sense” to realize that a regional strike or a minor hardware glitch requires a pivot that isn’t in its training set.
By removing the “fixed headcount” in favor of “fixed algorithms,” companies risk losing the very thing that makes them survive a crisis: human intuition. The goal of 2026 shouldn’t be to replace the workforce, but to use AI to handle the “highly repeatable tasks” so humans can focus on the 20% of outliers that cause 80% of the failures.
Final Thought: Your Role in the Collision
We are not entering an era where machines replace us; we are entering an era where our value is defined by how well we partner with them to manage chaos. If your business is “pure software,” you are at risk of the SaaSpocalypse. But if you can tie your digital intelligence to a physical “thick” execution, you aren’t just surviving the AI wave—you’re the one riding it.
The next time you see a “lights-out” warehouse demo, remember the protein powder. The future isn’t a perfect simulation; it’s a messy, physical reality, orchestrated by a thin layer of very smart code.
