I recently read an article celebrating how smartphones and Adobe tools are making everyone a creator. It was upbeat, confident, and polished. The kind of article that usually invites excitement or at least curiosity. Yet my reaction was unexpectedly flat.
There was no irritation, no urge to argue, not even skepticism. Just a mild sense of distance. I understood the claims. I recognized the logic. And still, nothing moved.
This lack of reaction surprised me more than disagreement would have. In the past, I might have felt admiration for such technological optimism. Or perhaps resistance. But this time, neither appeared.
The reason was not ignorance of the tools being praised. Quite the opposite. The reason was familiarity. Long familiarity, built through years of working inside software ecosystems that promise creativity while quietly shaping its boundaries.
That distance, I began to realize, was the starting point of a deeper reflection about AI, platforms, and the future of digital work.
Living Inside Software, Not Just Using It
For many people, Adobe and Microsoft are tools. For others, they are gateways. For those working in enterprise environments, they are closer to habitats.
Using systems like AEM makes this distinction very clear. Creativity does not begin with expression. It begins with permissions. With workflows. With templates. With governance rules that exist for good reasons, yet still impose structure long before meaning appears.
Over time, the software ceases to feel like an instrument and starts to feel like an environment one must enter. Once inside, movement is possible, but only along prescribed paths.
The famous periodic table of Adobe products captures this feeling perfectly. It looks playful, colorful, even celebratory. But for users, it also represents obligation. Each square is a function, a license, a boundary. Not all are needed, but many are unavoidable.
This does not make the software bad. In fact, its technical quality is part of the problem. The tools work well. They are reliable. They become embedded in professional life.
And precisely because of that, enthusiasm slowly fades into resignation.
Subscription as a Form of Dependence
Much of this feeling is tied to the subscription model, though not in the simplistic way it is often criticized.
The issue is not cost alone, although cost matters. The deeper issue is the shift in how continuity is experienced. With subscriptions, access is temporary by design. One is always renting stability.
This subtly changes the relationship between user and tool. Learning no longer feels like an investment in something that will remain. It feels like adaptation to a service that may evolve, reprice, or restructure at any moment.
The same feeling appears across Microsoft products. Word, PowerPoint, Teams, SharePoint, all are deeply capable. All are deeply embedded. And all reinforce the sense that professional work now takes place inside managed ecosystems rather than on neutral ground.
Again, this is not hostility toward these companies. It is recognition of their success. They have become indispensable.
But indispensability has an emotional cost. It flattens excitement. It replaces curiosity with maintenance.
The Strange Discomfort with Plug In AI
When AI began to enter these ecosystems, the initial response from platform vendors was predictable. AI would be added as a feature. A helper. A copilot. A generator.
From a business perspective, this makes sense. Existing users remain inside the same products. The learning curve stays shallow. Value propositions are easy to explain.
Yet something about this approach feels misplaced.
When AI is framed as a plug in, it is immediately reduced to utility. It becomes about speed, efficiency, and convenience. Write faster. Design faster. Summarize faster.
These are not trivial improvements. They are genuinely useful. But they are also limiting.
They quietly suggest that AI exists to serve existing workflows, rather than challenge them. That it is an accessory, not a foundation.
For those who sense AI as something more fundamental, this framing feels unsatisfying. Even misleading.
Productivity as a Narrow Lens
The dominant language around AI today is productivity. Increase output. Reduce effort. Optimize workflows.
This language is familiar, comfortable, and deeply aligned with corporate logic. It is also insufficient.
Productivity assumes that tasks are fixed, goals are clear, and outputs are already defined. AI simply accelerates what we were going to do anyway.
But the most interesting effect of AI is not acceleration. It is reconfiguration.
AI changes how ideas take shape. How drafts emerge. How thinking unfolds through dialogue. These changes are not easily measured in productivity metrics.
When AI is reduced to productivity, its deeper impact is postponed. We get faster versions of old habits instead of new ways of working.
This is why plug in AI feels like a containment strategy. It allows organizations to adopt AI without questioning the structures that define work itself.
AI as Infrastructure, Not Feature
A more convincing way to think about AI is as infrastructure.
Electricity was not an enhancement of candles. Writing was not an upgrade to memory. The internet was not a faster fax machine.
Each of these altered the conditions under which human activity took place. They changed expectations, rhythms, and forms of coordination.
AI belongs in this category.
It reshapes how language functions as an interface. How questions become actions. How drafts become conversations.
Seen this way, AI should not sit inside software. Software should sit on top of AI.
This inversion is subtle but important. It moves AI from helper to substrate. From feature to condition.
Once this shift is recognized, the obsession with productivity begins to feel oddly small.
The Unrealistic Dream That Will Not Disappear
There is a joke I sometimes make, half seriously, half defensively. That my dream is a digital world without Microsoft and Adobe.
The laughter that follows is usually polite. Everyone knows it is unrealistic.
Yet the dream persists.
It is not about eliminating companies. It is about eliminating compulsory entry points. About a world where creativity does not require entering corporate buildings, accepting licenses, or aligning with platforms before thought can begin.
Experienced users tend to have this dream more often than beginners. The more deeply one lives inside software ecosystems, the more one imagines what it would feel like to step outside them.
This is not rebellion. It is fatigue.
And it is one of the reasons AI inspires quiet hope.
Not Replacement, But Dissolution
It is important to be precise here. The hope is not that AI companies will replace Big Tech. That would simply recreate the same structure with new names.
The deeper hope is that no single layer becomes dominant enough to define everything.
Winner takes all logic has shaped the digital world for decades. Platforms grow, attract users, centralize value, and eventually become unavoidable.
AI has the potential to weaken this pattern, not by force, but by making centralization less necessary.
When intelligence can be distributed, localized, and personalized, scale loses some of its advantage. When interfaces become conversational rather than graphical, the importance of owning the interface diminishes.
This does not guarantee fairness or plurality. But it opens the possibility.
Can Intelligence Exist Without Platforms
The central question then becomes whether intelligence itself can be de-platformed.
This is not a technical question alone. It is social, economic, and ethical.
AI can just as easily reinforce centralization. Compute resources, data access, and model training already show signs of consolidation. The risk is real.
But unlike previous technologies, AI also lowers certain barriers. It allows individuals to work with capabilities that once required institutions. It supports thinking without formatting. Drafting without tools. Exploration without software overhead.
These small shifts matter. They create cracks in the assumption that all meaningful work must pass through platforms.
The outcome is not predetermined. But the possibility exists.
Creativity After the Tool Era
If AI continues to mature as infrastructure rather than feature, creativity may undergo a subtle transformation.
The center of gravity may move away from software mastery toward articulation. From tool expertise toward clarity of intent.
This does not diminish craft. It changes where craft resides.
Writing becomes thinking again, not formatting. Design becomes sense making, not menu navigation. Collaboration becomes dialogue, not file exchange.
In such a world, tools still exist. But they no longer dominate attention. They recede into the background.
That backgrounding may be the most important change of all.
Hope Without Hype
Returning to that initial flat reaction, it now feels less like indifference and more like clarity.
Excitement fades when one has seen enough cycles of promise and consolidation. What remains is a quieter form of hope.
Not the hope that AI will save us. But the hope that it will loosen certain assumptions. That it will weaken the idea that creativity must rent its tools, that intelligence must belong to platforms, that winners must always take all.
This hope is fragile. It depends on choices, norms, and restraint.
But it is real.
And perhaps that is enough to justify optimism, not as excitement, but as orientation.
AI does not guarantee a better world. But it makes a different one imaginable.
And sometimes, imagination is the most important infrastructure of all.
Image: Stockcake
Tom’s Blog 