Lately I’ve been thinking less about what AI can do and more about how we can tap into an expanded use of AI that can acquire more knowledge to help solve real problems.
Most conversations about AI in business focus on speed, automation, or replacement. That framing feels incomplete to me. Businesses rarely fail because they lack tools. They struggle because knowledge is scattered across people, projects, systems, and time.
So instead of asking what AI should automate, I started asking something simpler.
What layers of knowledge actually exist when people are working on real problems, and how could AI help us access more of those layers at once?
Businesses are always pulling from multiple layers of knowledge
When someone is trying to solve a business problem, they are never working from a single source of truth.
They are pulling from general knowledge, past experience, what they are seeing others struggle with, what their team knows, what their systems reveal, and what just feels off. Most of this never lives in one place.
That is why so much work is duplicated and why so many people feel like they are reinventing the wheel without realizing it.
Here is how I currently think about the layers of knowledge involved.
1. Universal knowledge
This is the knowledge that exists whether we have discovered it or not. Physics, undiscovered patterns, unknown relationships.
Neither humans nor AI really operate here directly. From a business perspective, this layer mostly serves as a reminder that there is always more to learn.
2. Global knowledge
This is knowledge discovered and shared by humanity and, increasingly, by AI systems.
Science, technology, economics, history, best practices.
Modern AI is extremely strong here. This is where it shines today. But this knowledge is general. It does not know your situation.
3. Social graph knowledge
This is one of the most underused layers.
Every social graph has its own knowledge. Founders. Small business owners. Creators. Nonprofits. Operators in a niche industry.
What keeps showing up
What people keep struggling with
What workarounds are common
What feels wrong before data confirms it
This layer does not diagnose or prescribe. It escalates attention.
It says this problem exists, persists, and costs time or money. Someone should be looking at this.
That alone is powerful.
4. Unit knowledge
This is where coordination begins.
A unit is a collection of entities working toward a broad problem.
That unit could be a company, a nonprofit, a solo founder, or even a single human coordinating AI agents.
This layer includes clarity around what problem is being pursued, why it matters, and what success looks like.
Without this layer, effort scatters.
5. Project knowledge
Projects are how progress happens.
Plans, experiments, decisions, iterations, results.
Projects are temporary, but the knowledge they produce should not be. Too often, lessons disappear when projects end.
6. Entity knowledge
Every entity has knowledge.
A human.
An AI agent.
A system.
Each has different capabilities, limits, memory, and perspective.
Scale does not come from one entity doing everything. It comes from coordinating entities well.
7. Individual human knowledge
This layer is distinct and always will be.
Judgment.
Values.
Risk tolerance.
Meaning.
AI can support this layer, but it cannot replace it. This is where decisions are made about what problems are worth solving in the first place.
Why this matters to me
Most people and businesses operate as if they are alone.
They solve the same problems.
They rebuild the same solutions.
They waste time not knowing what others already know.
Now imagine hundreds or thousands of entities operating with shared background awareness instead of isolation.
Not collaborating on the same solution.
Not losing independence.
Not sharing everything.
Just aware of what problems exist, which ones are growing, and which ones are already crowded.
That does not reduce creativity. It focuses it.
What I think AI should really be doing
AI does not need to decide what we build.
Its real value is in helping us see more clearly.
Connecting knowledge across layers
Preserving context over time
Surfacing patterns no single person can see
Helping people decide where to apply effort
The future of business is not about having more information. It is about knowing where to look, what matters, and how to coordinate without wasting energy.
That is the direction I keep coming back to.
