which layer of your business is AI-proof?
Last year the question everyone asked was what is your AI strategy.
This year the question is what makes it last.
Boards ask it. Customers ask it. Hiring panels ask it.
Your team asks it in the all-hands you have not yet scheduled. The question travels.
The new word is durability. It is the word capital uses when AI starts eating margins. It is the word your customers use when they decide to renew or churn. It is the word your team uses when they decide whether to stay or leave.
Underneath the word is a simple question.
Three years from now, when AI is much better, what part of your business is still worth what it is worth today?
That question does not get answered by saying “we use AI.”
It gets answered by saying which part of your business AI makes stronger, and which part AI makes obsolete.
Capability is cheap. Position matters.
Every company sits inside a stack.
At the bottom: physical things. Chips, energy, capital.
In the middle: the doing. Code running. Workflows completing.
At the top: the meaning. Trust, brand, what only you know about your customers.
That slice you sit in is your layer. The full set of them has a name: Economic Abstraction Layers. It is how investors think about durability in any business, and it is the lens you have to learn to think in too.
For most of the last decade, every slice was worth something. AI changed that.
AI is great at the middle. It cannot touch the top. It does not change the bottom.
So the durability question becomes simple.
What layer do you own?
And what does AI do to it?
Economic Abstraction Layers, plain language
Substrate Layer. Chips, energy, the physical world. Oe.g.TVIDIA., TSMC.
Infra Layer. The cloud and the pipes. e.g. AWS, Cloudflare.
Execution Layer. The doing. Code running. Tasks completing. e.g. Salesforce, Stripe.
Collaboration Layer. How teams work together. e.g. Slack, Notion, Linear.
Coordination Layer. How systems talk to each other. e.g. Plaid, Twilio.
Translation Layer. Turning capability into something usable. e.g. Apple, Spotify.
Brand Layer. Who the company is. Why customers love it. e.g. Patagonia, Liquid Death.
Identity Layer. Knowing who the user is. KYC, login. e.g. Okta, Plaid Identity, Onfido.
Context Layer. The data and signal nobody else has. e.g. Bloomberg, Epic.
Assumptions Layer. The thesis. The worldview the company runs on. e.g. Tesla, OpenAI.
Governance Layer. Trust, audit, accountability. e.g. Visa, bond rating agencies.
What AI does to each layer
Three buckets:
AI eats Execution and Translation layers.
The doing layer and the form-it-takes layer. Most of what consulting firms, BPO shops, and entry-level analysts do for a living. Margins compress to zero.
AI consolidates Substrate and Infra layers.
The physical and the cloud. AI workloads need scale, and only the largest players win there.
AI depends on Context, Coordination, Governance, Identity, and Assumptions layers.
These five get more valuable as AI improves.
Because AI needs data to reason against (Context). It needs to plug into the rest of your systems (Coordination). It needs to clear regulatory bars before it can ship (Governance). It needs a brand customers recognize and trust (Identity). And it inherits the worldview of whoever trained or deployed it (Assumptions).
Five durable layers. Five commoditizing or consolidating layers.
The companies that compound through the next decade will be operating in the most durable five. This is also true one level down. The teams and roles that compound are the ones operating in the durable layers. If your work mostly lives in execution or translation, AI is coming for your output. If your work lives in context, coordination, trust, brand, or judgment, AI makes you more valuable, not less.
The same lens that decides company durability decides team durability and individual durability.
The diagnostic
Take this into your next board meeting, pitch, or strategy review.
What gets stronger when AI gets better at everything else?
That is the test.
The answer to it is the answer to one question.
What layer do you own?
Whatever you can name is the layer worth building around.
If you cannot answer, AI eats your business over the next three years.
If you can (the trust you have built, the data you own, the brand customers love, the workflows other companies plug into), build around the answer.
The narrative underneath the question
Investors already think in layers.
So do customers when they renew. So do executives when they decide which team gets the budget.
Operators usually do not. That is the gap.
The work is two moves:
First, name the layer. Not “we use AI.” That is a category, not a position. Name the specific Economic Abstraction Layer that gets stronger as AI gets better. Second, build the narrative around the layer.
The narrative is what makes the layer real to anyone who does not already think in layers. It tells customers why they renew. It tells investors why the moat will hold three years out. It tells your team what they are building toward. It tells executives why your work deserves the budget. Without the narrative, the layer stays a private fact. With it, the layer becomes the lens every stakeholder uses to understand what your work is worth.
In your next high-stakes conversation about your work, whether that is a board update, customer pitch, all-hands, kickoff, performance review, or team Q&A, lead with the layer. Then build the narrative around it.
That is the conversation capital, customers, peers, and your team are already trying to have with you.
Now you have a way to give it to them.