Artificial intelligence now sits inside almost every tool you open, from search engines and office apps to browsers, phones, and creative software. Every update brings new assistants, copilots, and generators, each one promising to change how work gets done. On paper, adoption looks high. Millions of users already have these features available, often switched on by default, waiting inside menus that most people rarely explore. Yet actual behaviour moves more slowly. Many users still write documents line by line, search the web the same way they did years ago, and complete tasks manually, even when the software suggests another option.
The disconnect between availability and usage is a well-known phenomenon in technology circles. It mirrors earlier waves of innovation, from graphical user interfaces to cloud computing. But the scale of AI deployment is unprecedented. According to a 2024 Microsoft Work Trend Index, 66% of leaders say they wouldn't hire someone without AI skills. Yet the same research shows that only a minority of employees feel properly trained to use AI at work. This training gap persists even as tools are pushed into everyday use. The result is a workforce that has access to powerful capabilities but lacks the confidence or know-how to integrate them into daily routines.
Feature Overload Makes Modern Software Harder to Use
Modern apps are not struggling because they lack capability. They struggle because every update adds another layer on top of what was already there. AI did not replace old interfaces; it stacked on top of them. Users now face more options, more panels, and more assistants than before. Open almost any tool today and the pattern looks familiar: office software with built-in copilots and sidebars, design tools filled with generators, templates, and prompts, productivity apps with chatbots inside every menu, and platforms that expect users to learn through guides similar to employee training. When the interface becomes crowded, people stop experimenting and return to what they already know.
This phenomenon is often called feature fatigue. A 2023 study by the Nielsen Norman Group found that users abandon tasks when interfaces exceed a certain threshold of visible options. The cognitive load of evaluating which feature to use, especially when new AI-driven capabilities are buried under unfamiliar icons or tucked inside sidebars, can outweigh the perceived benefits. So instead of learning how to use a copilot to summarise a document, many users simply continue typing summaries by hand. The irony is that the very tools designed to save time often introduce decision paralysis that wastes it.
Historical Parallels and the Human Resistance to Change
This pattern repeats across every major technology shift. When word processors replaced typewriters, many writers refused to adopt them because they feared losing the tactile feedback of keys. When spreadsheets replaced ledger books, accountants initially stuck with paper. In each case, the new tool eventually won, but only after generations of users aged out and new ones grew up with the technology. Today's AI adoption faces a similar inertia, but the pace of change is much faster. Software updates now roll out weekly, not yearly. Users are expected to continuously update their mental models of how applications work, a demand that clashes with the way human cognition prefers stable routines.
Behavioural economics offers another lens. The status quo bias means people favour existing habits over new alternatives, even when the alternatives are objectively better. This bias is especially strong when the new option requires upfront learning. The brain perceives the learning cost as a loss, and loss aversion makes it harder to switch. A 2022 study from Carnegie Mellon University found that users are 70% more likely to reject a new workflow if it requires more than five minutes of initial exploration. That is a frightening threshold for software vendors who release complex AI features that demand time to understand.
The Missing Piece: In-App Learning and Digital Adoption Platforms
Software vendors are not moving slowly. New AI features appear in updates almost every week, added to tools people already use for writing, coding, design, search, and communication. Access is no longer the barrier. What is missing is the moment when the user actually learns where the new feature fits into their existing workflow. Most software still expects people to figure that out on their own. This realisation has given rise to a category of tools known as digital adoption platforms (DAPs). Companies like WalkMe, Pendo, and Appcues now focus on teaching features within the application rather than sending users to separate documentation or training portals.
WalkMe Learning Arc, for instance, overlays step-by-step walkthroughs directly inside the interface. Instead of reading a manual, the user sees a highlighted area on the screen with a suggestion: 'Try asking the copilot to summarise this document.' The action is contextual, immediate, and requires no external search. This approach reduces the cognitive gap between knowing a feature exists and knowing how to apply it. Early data from DAP vendors suggests that organisations using in-app learning see a 40% higher adoption rate for new features compared to those relying on traditional training sessions.
Yet even the best in-app guidance cannot overcome all barriers. Many enterprise tools are still built on decades-old interfaces with layers of legacy options. AI features are grafted onto systems that were never designed for them. This creates a user experience that is disjointed and confusing. A 2024 report from Gartner predicted that 60% of organisations will struggle to realise value from AI investments due to user adoption challenges. The report recommends focusing on friction reduction and contextual training rather than simply adding more AI capabilities.
People Don't Resist AI; They Resist Changing How They Work
Most users are not against artificial intelligence. What they resist is changing the way they already know how to work. Once a routine feels reliable, people repeat it without thinking, even when the software offers a faster method. Habit becomes the default. This helps explain why the gap is growing between AI availability and real capability. Many employees are learning on their own while job requirements move closer to the skill sets now associated with future roles rather than traditional positions. The gap between innovation and adoption is mostly human, not technical, which is why the next shift in AI will not come from better models alone.
Consider the example of email management. AI can now summarise threads, suggest replies, and even automatically sort messages into priority folders. Yet many users still manually read every email and type replies from scratch. They have done so for years, and the muscle memory is strong. Only when a tool offers a truly frictionless alternative do habits shift. For instance, Gmail's Smart Reply feature saw high adoption because it appears as a simple clickable button at the bottom of the message, requiring zero learning. The same principle applies to AI copilots in Microsoft Office: if the user must locate a sidebar or remember a voice command, the barrier is often too high.
The Next Wave of AI Will Focus on Teaching, Not Just Automating
The next phase of AI development is starting to move away from adding more features and toward helping users understand the ones already there. Instead of expecting people to read guides or watch tutorials like it's 2015, newer tools are beginning to guide actions directly within the interface. Copilots that recommend the next command, walkthroughs that appear in the middle of a workflow, and interfaces that adapt to how the user works are becoming more common across productivity, design, and development software.
Some forward-thinking companies are embedding AI tutors directly into their applications. For example, Adobe's Photoshop now includes an in-app "tutorial" mode that highlights AI features like generative fill, not just in a separate panel but by dimming the rest of the interface and overlaying step-by-step hints. Similarly, Figma's AI plugin suggestions appear only when the user's behaviour indicates they might need them, such as after repeatedly performing a manual task like resizing multiple objects. This shift from "here's a new feature" to "here's what you can do right now" is critical.
This approach also reduces the need for formal training programmes. Traditional employee training, like workshops or e-learning modules, suffers from a transfer problem: learners retain information poorly when they cannot apply it immediately. In-app learning solves this by providing "just-in-time" guidance. Digital adoption platforms are now being integrated into enterprise software suites to provide this. When an employee opens a CRM tool for the first time, they see a popup that says, "Want to automate logging calls? Click here." The learning happens in the context of real work, making it stickier.
The implications for software design are profound. The tools that stand out will not be the ones with the longest feature lists, but the ones people can actually understand without stopping their work to figure them out. This places a premium on user research and iterative onboarding. Companies that invest in reducing cognitive friction will see higher adoption and better return on their AI investments. Meanwhile, those that continue to release features without contextual guidance will watch their users remain stuck in 2015 workflows, no matter how many copilots they add.
Source: TNW | Insights News