software is dead. context is the product.
So if everyone can build software, what becomes valuable? I think the answer is context.
Listendownload0:000:00
Software Is Dead. Context Is the Product.
Software used to be the competitive advantage. Now, the ability to build it is becoming a commodity.
AI is making code abundant. Almost anyone can turn an idea into a working application. What once took months can sometimes take hours.
So if everyone can build software, what becomes valuable?
I think the answer is context.
Who you are. What you own. Where you've been. Who you trust. The experiences you've had and the relationships you've built.
Software can be replicated. Meaningful context is much harder to reproduce.
And I think this changes how we should think about products.
From software to context
Most software today still expects us to do the work. We open applications, provide information, navigate interfaces, and repeat ourselves across different services.
But what happens when software already understands us?
We're beginning to see this with personal AI agents. Instead of interacting with a generic assistant, imagine an agent that understands your work, your relationships, your preferences, and the decisions you've made over time.
An agent that remembers what matters to you, understands what you're trying to accomplish, and can act on your behalf.
The interesting part isn't necessarily the underlying AI model. Thousands of applications can access similar models.
The difference is what those models know about you.
Your history, your intentions, your relationships, and the context accumulated through everyday interactions.
This is where I think personal software becomes interesting again. Not another productivity app with AI features, but software that becomes increasingly useful as it develops a deeper understanding of you.
The interface might be a conversation, a notification, or an action happening in the background. You might not even need to open an app.
The product is no longer just the software. It's the relationship between intelligence and your context.
Your data is the advantage
The same idea extends to companies.
For years, businesses have accumulated enormous amounts of proprietary data. Customer relationships, internal processes, operational knowledge, historical decisions, and information scattered across hundreds of tools.
Most of this data remains disconnected or underutilized.
Now imagine an AI agent that understands how a particular company operates. Not just because it has access to documents, but because it can interpret the company's history, workflows, customers, and institutional knowledge.
Two companies could use the exact same AI model and get dramatically different results because of the context available to each.
This changes where competitive advantages might come from.
The underlying intelligence becomes increasingly accessible. But the context that makes it useful can be unique.
Of course, having more data doesn't automatically mean having better intelligence. The information needs to be relevant, reliable, accessible, and used with the right permissions.
But I think we're moving toward a world where the value isn't just in building the software, but in the context that makes the software useful.
Intelligence moves into the physical world
This becomes even more interesting when we look beyond screens.
Robotics is one example.
A general-purpose robot might understand how to perform a task. But a robot operating inside a warehouse, hospital, or factory needs much more than general intelligence.
It needs to understand its environment, the equipment around it, the people it works with, and the specific procedures of that organization.
That context could come from sensors, operational data, human demonstrations, historical records, and experience accumulated over time.
The intelligence might be general-purpose, but its usefulness depends on understanding the specific world in which it operates.
And this isn't limited to robots.
We've spent decades designing digital experiences that live inside screens. But the next generation of products might not look like applications at all.
They might look like the things we already own and use.
A jersey that remembers every match you've attended. A diploma that can verify its own history. A wedding ring that carries decades of memories.
Not because everything needs AI, but because physical objects can become gateways to experiences that persist beyond the moment of purchase.
Imagine buying a product and having that purchase mark the beginning of a relationship rather than the end of a transaction.
The product has an identity. It can accumulate history, connect to experiences, and become more meaningful over time.
This is something we're experimenting with at kode, using fashion as a starting point to explore connected physical products and digital identity.
But the idea extends far beyond clothing.
The world is the interface
I think we're witnessing a shift similar to previous generations of computing.
The web connected information. Mobile connected people. AI is making intelligence accessible. And now that intelligence has the potential to extend into the physical world.
Personal agents, company-specific intelligence, robotics, and connected products might seem like different categories, but I think they share a common foundation.
Context is what makes general intelligence personally or situationally useful.
And as intelligence becomes more accessible, the interfaces we use to interact with it may become less important than the context behind them.
That doesn't mean applications disappear tomorrow, or that every object needs to become intelligent.
It means we might need to rethink what we're actually building.
Instead of asking what software we can create, perhaps the better question is what relationships, knowledge, and experiences we can make possible.
Because when software becomes abundant, knowing what matters becomes the difficult part.
And maybe that's where the next generation of products begins.
Software is dead. Context is the product. The world is the interface.