The design process we use today was built for a different era.

For the past couple of decades, most product teams have followed some version of the same playbook. Research, synthesis, personas, journey maps, prototypes, testing. Over time it became so familiar that many organizations started treating it almost like a formula.

But if we’re honest, the results haven’t always been as transformative as the process promised.

People still struggle with basic digital experiences every day. You open a banking app to transfer money and end up digging through five menus. You log into a healthcare portal only to realize you need a different account to see your test results. You try to cancel a subscription and somehow end up stuck in a chatbot loop.

The process may be thoughtful. The outcomes often aren’t. Part of the reason is that the systems we’re designing today are very different from the software those methods were originally built for.

When products stop behaving predictably

AI is quickly becoming part of everyday products and services. Systems now recommend, predict, flag unusual activity, automate decisions, and adapt based on the data they receive.

That changes what the “experience” actually is.

Take credit card fraud detection. When a suspicious charge appears, the bank often blocks it before you even notice. You receive a quick notification asking if the purchase was legitimate, tap yes or no, and move on with your day.

From the customer’s perspective, the experience feels simple. But the real experience wasn’t the notification. It was the system behind it analyzing behavior, comparing patterns across millions of transactions, and deciding whether to intervene.

We’re seeing the same pattern across industries. Logistics platforms reroute shipments before delays happen. Streaming services adjust recommendations as viewing patterns evolve. Some airlines automatically rebook passengers when disruptions occur and simply send a new boarding pass.

In many cases, the most important part of the experience happens before the interface even appears. And that raises a bigger shift that the design industry hasn’t fully grappled with yet.

The best experience might remove the interaction

For years, digital products have been optimized for engagement. More clicks, more sessions, more time inside the product. Entire industries were built around increasing interaction. But most interactions people have with technology aren’t things they actually want to do.

Nobody wakes up excited to check whether a payment cleared, refresh a shipping page, confirm a flight change, or navigate a customer service flow. These are small pieces of digital overhead that people have simply learned to tolerate.

As systems become more intelligent, many of those moments start to disappear. A shipment reroutes automatically. A suspicious charge is handled before it becomes a problem. A meeting is scheduled without ten back-and-forth emails.

In that world, the best experience isn’t the one with the most elegant interface. It’s the one that quietly removes the need for the interaction in the first place.

Which means the future of experience design may not be about designing more interactions at all. It may be about designing systems that eliminate them.

Designing how systems behave

This is where the traditional design process begins to feel limited.

Personas and journey maps were incredibly useful tools for helping teams understand people and the situations they move through. But they were developed in a world where software behaved predictably. A user moved through a series of steps and the system responded in defined ways.

AI-driven systems don’t operate like that. They evolve as they learn from data. They adapt to context. They sometimes produce outcomes that even the teams building them didn’t explicitly program.

In those environments, the experience isn’t really a fixed journey anymore. It’s an ongoing interaction between a person and a system that continues to evolve.

That shifts the design challenge deeper into the system itself. Instead of focusing primarily on screens and flows, teams increasingly need to think about how systems behave. How they make decisions, how they respond when the data is uncertain, and when automation should step aside for human judgment.

Those questions rarely sit neatly inside one discipline. Designers, engineers, product leaders, and data scientists all shape how these systems operate and how people ultimately experience them.

Rethinking the process

None of this means the principles of human-centered design stop mattering. If anything, they matter more as systems grow more complex. But the artifacts and processes that shaped the last twenty years of design practice were built for predictable software.

The systems we’re building now are adaptive, data-driven, and constantly evolving. The opportunity isn’t simply using AI to make the existing process faster. It’s stepping back and asking what the design process should look like for systems that learn.