Post-Platform diversion: How Computing’s Disappearance Will Reshape Everything

Third Essay in our series about REVERSIBLE COMPUTING and it's interaction with MODEL CONTEXT PROTOCOL.

What does it look like if you're no longer shipping apps but simply weaving experiences?

There’s a moment, just before a thunderstorm rolls in when the air feels alive... charged. Not chaotic but alert; every breath is held because something’s about to shift.

That’s where we are with computing and the coming change is something a lot stronger than the oft played “faster chips, larger memory, more processing” song.

A world where the machine disappears altogether.

Thanks to the convergence of reversible computing and something called the Model-Context Protocol (MCP), we’re stepping into a future where logic isn’t hosted, it’s summoned, where data doesn’t live inside apps it flows through context and where software isn’t a product at all -  it’s a moment of meaning that assembles, executes, and dissolves.

This shift isn’t just technical. It will ripple through work, governance, markets, and identity; The idea of a “platform” (that thing we’ve been building everything on for the last two decades) begins to fade. And in its place, something far more fluid starts to emerge.

Let’s take a look at what happens when software stops staying put.

The Work After Apps

There will be entire industries around this new data usage system. In a world where logic is ephemeral nothing needs to be stored, systems don’t persist unless they’re being used and the kind of work that’s needed is different.

We’ll still need builders. But we won’t need as many platform engineers, dashboard designers, or people writing code to wrangle other code. Instead, we’ll need flowcraft.

Flowcrafters are conductors, not constructors. They don’t build static apps - they design experiences that can be assembled in real time, adapted on the fly, and reversed if needed.

Some will specialize in context engineering: helping companies translate messy business logic into structured, machine-readable context that can be queried by intelligent agents. These folks won’t build apps — they’ll codify culture, pricing logic, escalation policy, tone of voice. Not for UI menus, but for AI systems that interpret and act on it in real time.

Others will be orchestration studios: think choreography for logic. A logistics firm won’t build a shipment tracker. It’ll hire a flow studio to design a set of reversible micro-interactions: reroute a package, resolve a customs snag, update an ETA. Each flow runs briefly, contextually, and then disappears.

And in a world where logic vanishes after execution, trust becomes everything.

New firms will emerge to verify that a computation happened the way it was supposed to — without storing the computation itself. They’ll certify that a flow was built from known parts, used the right data, and ran under approved conditions. Footprints instead of logs.

There will be hardware players too, building reversible processors that sip power instead of burning it. The world of “smart” devices could shift to your watch as the hub with much simpler, lighter or even flexible devices for phone, tablet and laptop.

And then there’s semantics. Not everyone speaks the same context. So we’ll need semantic brokers, people or systems that translate between ontologies, align meaning, and help autonomous agents avoid misunderstandings, a friend and mentor I miss having time with used to describe talking to his father-in-law about cutting a 12 inch piece of would - my friend would say “cut 6 inches”, his father-in-law would say “leave 6 inches” - depending on how precise your project that difference might mean nothing or everything.

The shape of labor shifts. Developers give way to flow architects. Project managers become context strategists. Compliance officers evolve into MCP delegates. And we’ll need computational ethnographers — people who study how humans interact with momentary logic, so we can make it usable, humane, and meaningful.

None of these roles are science fiction. They’re just slightly sideways from the ones we already have — nudged by a different set of assumptions.

The Politics of Disappearance

Of course, this isn’t just an economic shift. It’s political. Deeply so.

Reversible computing dramatically reduces the energy needed to do work. Entire data centers become obsolete. Countries that adopt this model early can reduce their dependence on hyperscale infrastructure and redefine what energy sovereignty looks like. Fewer servers, more sovereignty.

But control won’t vanish — it will migrate. In the same way cloud platforms became gatekeepers, the orchestration layer will be the new battleground. Whoever controls the flow registries and the semantic scaffolding will control access to computation itself.

Identity also gets weird.

In systems that don’t persist, who you are becomes a question of context. Privacy law will have to evolve. Regulation will need to govern temporary access, delegated consent, and data that forgets — but still infers. 

Personal agents representing us will become a normal part of life - we’re already one step closer with OpenAI releasing memory capability across your chatGPT discussions, let’s take that a step further in a world where the cost of compute is virtually non-existent - we have the opportunity to create, manage and leverage a true digital twin that can engage with our environment and act on our behalf.   We own our data stories and choose who we’ll allow in to our digital minds.

There will be debates. Should ephemeral logic be reconstructable? Should governments have the right to audit flows after they dissolve? We may end up with regulated transparency zones — places where reversibility is the norm, but certain flows must be observed and retained.

Governance won’t die with the app. It’ll just become ambient, too.

The Economic Fallout

The old pricing models won’t hold up.

SaaS subscriptions? Gone. When software no longer persists, there’s nothing to subscribe to. Instead, you’ll pay per flow, per result, per moment of intelligence.

Compute costs will collapse. Logic will run on the edge, powered by local context and barely any energy. The constraint shifts from processing power to meaning — from bandwidth to trust.

Governments may step in. Logic-as-a-utility is no longer a metaphor. It’s quite possible that ephemeral flows will be metered, subsidized, or taxed. Especially when they’re used for public goods.

Cloud providers will have to adapt. They won’t charge for uptime anymore. Instead, they’ll offer curated context, certified execution environments, and semantic tooling.

And maybe most powerfully: compute becomes democratized. Because reversible logic runs cool and local, entire regions that were once left out of the digital economy may leapfrog straight into autonomy. No hyperscale cloud required.

Conclusion: The Platform Dissolves

This isn’t just a technical shift. It’s the slow(ish) dismantling of the assumptions we’ve built the last two decades on.

In the new model, we don’t store - we stream, we don’t host  - we assemble, we don’t ship apps - we define intent, and let flows form as needed.

The winners won’t be the ones with the biggest clouds. They’ll be the ones who define the clearest flows, build the most trustworthy semantics, and design systems that don’t just run — they listen.

We’re not building platforms anymore, we’re building the weather. We need a mindset shift  to think about software as a moment; a transformation that happens, leaves a trace, and then disappears.

The next operating system isn’t an OS - it’s a breeze, a gust of wind… Let’s learn how to read it and shape it - together.