Why Most Knowledge Workers Do Not Wear Shoes Yet
AI capability is here. Adoption lags because work still runs on friction. The opportunity is an execution layer with proof.
Benjamin Zander tells a very compelling story about the two salesmen who went down to Africa in the 1900s. They were sent down to find if there was any opportunity for selling shoes, and they wrote telegrams back to Manchester. And one of them wrote, “Situation hopeless. Stop. They don’t wear shoes.” And the other one wrote, “Glorious opportunity. They don’t have any shoes yet.”
That single word changes everything. The first person sees a lack as proof of impossibility. The second person sees a lack as proof of untapped demand.
Most knowledge workers today do not wear shoes yet.
Not because they’re unintelligent. Not because they’re lazy. Not because the tools aren’t here. The constraint is that modern work still runs on friction: humans acting as the glue between systems.
What “shoes” means in knowledge work
Shoes are leverage. They turn the same human energy into more distance, more stability, less injury, and more consistency.
For a knowledge worker, “shoes” isn’t another app. It’s an execution substrate that:
- starts from intent (what you want)
- runs work in a reliable environment (not vibes)
- produces artifacts (not just advice)
- proves what happened (not “it should work”)
In other words: a system that compresses the path from “I intend” to “It exists.”
Why we’re still barefoot
If AI is moving this fast, why aren’t most people already operating with this kind of leverage?
Because the surface area of modern work is still optimized for manual control.
- We’re trained on interfaces, not outcomes.
- SaaS fractured the workspace. Humans became the integration layer.
- Most “agents” stop at suggestions. The expensive part is end-to-end execution with proof.
- Trust breaks at the moment you need it most (money, customers, production).
So we stay barefoot: maximum attention burn for minimum output.
Capability is compounding. Comprehension is not.
There is another reason this moment feels chaotic: the pace of AI is outpacing humanity’s ability to integrate it into day-to-day work.
When the rate of change exceeds our ability to digest it, the winning move isn’t “learn every new model.”
The winning move is to adopt a stable abstraction that makes model improvements automatically translate into real outcomes.
That abstraction is an execution layer.
The execution layer is the adoption unlock
OutcomeDev exists because the bottleneck moved.
The differentiator is no longer who can demo intelligence. The differentiator is who can turn intent into repeatable execution, shipped artifacts, and proof (without chaining the human to a desk and a thousand clicks).
This is why “repos” matter: they’re the simplest durable container for memory, constraints, artifacts, and audit.
The next level: the agent workday
Once you have an execution layer, time itself becomes a vehicle for change.
You can run a complete operating loop in a single bounded runtime window:
- observe changes
- decide what matters
- execute actions
- produce proof
- update durable state
That’s the Agent Workday:
And if you can re-run the loop daily, you’ve built an operating system for work; not by writing a new product, but by running repeatable tasks:
“Most knowledge workers don’t wear shoes yet because they’re still the glue between tools.” Brighton Mlambo
The glorious opportunity
The Zander story isn’t really about shoes. It’s about how you interpret the gap between capability and adoption.
You can look at the modern workspace and say: “Hopeless. People don’t work like this.”
Or you can look at it and say: “Glorious opportunity. People don’t work like this yet.”
If you could cut your screen time in half but double what you ship, would you?
That question is why OutcomeDev exists.