Why The Interface Is Evolving From Navigation To Intent

Dr. Son Nguyen is the cofounder & CEO of Neurond AI, a company providing world-class artificial intelligence and data science services.

getty

Over the past several decades, companies have added tool after tool to manage every corner of their business, from customer data to internal approvals. While the interfaces got cleaner and the features got smarter, the work somehow got more complicated.

Asana reported that employees spend up to 60% of their workday on “work to work" like navigating software, switching between apps, updating statuses and filling out forms. Why does this keep happening? How is AI changing this situation?

The Revolution Of UI

At its core, a user interface is a translation system that helps turn human intent into machine action.

Historically, this translation was necessary because computers were “blind” to intent. They couldn’t infer what you wanted. And every generation of computing has tried to close that gap. None of them fully succeeded, but each one got closer.

The mainframe forced users to fully adapt to the machine. You encoded your intent physically, one punch card at a time.

Things got faster with the command-line interface (CLI), but not necessarily easier. You could issue commands quickly, but even a small typo could break the interaction.

The graphical user interface (GUI) was the first real breakthrough, replacing memorized commands with visual interaction through files, folders and icons.

Then came touch interfaces, pushing this further, letting users directly manipulate digital objects with their fingers.

Each step made computing more natural, more human.

Why Interfaces Are Becoming A Burden for Companies

If you look at a modern enterprise workstation, much of the modern workday is still dominated by UI labor. Employees aren’t just thinking or making decisions; they’re operating software. They act as human middleware, copying data from a PDF into a spreadsheet, pasting information from your CRM into an email and reformatting it. None of this is real work. ​

Specialized SaaS tools compounded this problem, as each one introduced more tabs. As of last year, six in 10 IT leaders say their organization introduces a new SaaS tool every month, according to a Nintex survey. Every new “solution” adds another interface to learn, another login to manage and another workflow to master.

We're reaching the limit of how much UI a human can effectively manage.

The Fifth Era Of UI: Intent-First Computing

Entering the AI era, the interface seems to be disappearing, and the way we interact with software is fundamentally changing.

Instead of telling computers how to do something step by step, you let them know what you want. The system figures out the steps, executes them and returns a result for review.​

This is what intent-first computing means. This shift didn’t happen earlier since the technology simply wasn’t ready. Today, LLMs can understand natural language with remarkable accuracy. AI agents can also autonomously handle complex workflows. The cost of computing has dropped enough to make these systems practical at scale.

For a computer to truly disappear into the background, it must understand more than just your words. It has to master three types of intent:

• Explicit Intent: What you ask for.

• Implicit Intent: Unstated expectations, like formatting or using the latest data.

• Ambient Intent: Contextual knowledge such as your preferences and past behavior.

So, is UI disappearing? No, but is being reorganized around a completely different job. Traditional UI focused on navigation through menus and workflows. The new one is built around four layers that work together in the background:

• The intent surface uses simple inputs like text or voice prompts to interact with the system.

• The inference layer uses AI agents to interpret goals, plan the steps and execute them.

• The review layer helps evaluate the result. You approve, edit, correct or ask for a different version.

• The memory layer makes the system smarter over time by learning your preferences, context and history.

This also introduces a bigger change in how software behaves. Traditional software is deterministic: you click “Button A” to get “Result B,” with no exception.

AI systems are probabilistic. The same request can return slightly different results depending on context or phrasing. This is why the “review layer” is so important. Now, users' value isn’t how fast they can navigate but how well they judge the quality of the outcome. ​

What This Means for Companies and Engineers

For many businesses, this ship is already here. Optimizing screens and flows is no longer enough.

The first thing to let go of is the instinct to build complex interfaces. AI is addressing the problem of long onboarding flows, multi-step dashboards and elaborate navigation systems.

However, the new design priority is simple to say and difficult to execute. Your systems must understand what people want, execute it reliably and make it easy to review and trust the result. Intent, execution, oversight. That is the new stack.

Noticeably, adding a thin layer of AI to existing products won’t be a durable strategy. Many of these experiences are easy to replicate and offer limited differentiation. The harder and more important question to ask is: What would this product look like if it were built for intent from the start?

Real, lasting value in the AI era comes from three things: your data, your workflows and the trust you've built with users who rely on your outputs to make decisions.

For engineers, the day-to-day experience will also change. The trend is toward more integrated systems. You will spend more time architecting the logic that allows AI to make correct decisions.

Now, the advantage lies in owning the data and owning the workflow. If you control the “brain” and the “memory” of the system, you control the value.

Conclusion ​

UI isn’t disappearing overnight. Buttons, menus and dropdowns existed because computers couldn't understand people. So we built translation layers to bridge the gap. For decades, those layers served us well. But a workaround is still a workaround.​

​AI is now reducing the need for those layers by allowing users to simply express what they want and evaluate the results.

For companies, this is a moment to rethink where effort is spent. For engineers, it’s a shift in what it means to build great software. And for users, it marks the beginning of a very different relationship with technology. ​

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?