The Illusion of Demand: Why Most People Don’t Want More AI, But Instead Need It to Do the Boring Stuff
Many businesses operate under a pervasive assumption: that the public is actively seeking and anticipating an influx of new artificial intelligence features, products, and workflows. This optimistic outlook posits that AI will seamlessly integrate into daily life, rendering existing methods obsolete and correcting inefficiencies. However, emerging data and user behavior suggest a starkly different reality. The prevailing sentiment indicates that a significant portion of the population does not necessarily desire more AI, particularly not in the manner currently envisioned by many AI developers and industry leaders. This disconnect is evidenced by the notable low adoption and retention rates of numerous AI features, often introduced at substantial development costs and carrying the risk of reputational damage.
The Unmet Promise: AI as a Feature, Not a Solution
The core of the issue, as highlighted by industry analysis and user feedback, lies in the fundamental misconception that "AI" itself is a compelling value proposition. Senior leadership within many organizations struggles to articulate the tangible benefits of integrating AI, often viewing it as a technological advancement rather than a solution to a specific problem. This approach frequently results in AI features being tacked onto existing systems as separate tools. Consequently, employees are often pulled away from their established and optimized routines to learn and operate these new, sometimes clunky, interfaces. This disruption can negate any potential efficiency gains, leading to frustration and reduced productivity.
Artificial intelligence, in its current widespread implementation, often functions as an amplifier of pre-existing organizational shortcomings. Issues such as poor data quality, flawed decision-making processes, and accumulated technical debt become more apparent and problematic when subjected to AI analysis. Rather than magically resolving years of accumulated workarounds, cultural inconsistencies, or internal politics, AI can inadvertently expose these vulnerabilities. The inconsistencies and conflicting priorities are laid bare, and users are left to navigate the resulting complexities, often without adequate support or clear guidance.
The fragmented nature of modern work environments, characterized by numerous disconnected systems, exacerbates this problem. The introduction of a new AI tool often adds yet another system to the user’s repertoire, requiring them to switch contexts and manage yet another layer of complexity. This can paradoxically lead to increased workload and a diminished sense of reward for the tasks performed.
Furthermore, individuals are increasingly aware of the significant effort and time required to identify and rectify errors generated by AI, commonly referred to as "hallucinations." While generating content or performing tasks with AI might appear easier than starting from scratch, this perceived ease comes with a hidden cost: the labor involved in verification and correction. This burden can outweigh the initial time savings, leading to a net increase in the overall effort required.
The Psychology of Resistance: Fear, Anxiety, and Uninvited Change
A significant factor contributing to the low adoption of AI features is the lack of user agency. For many, AI arrives as an imposed change, dictated by organizational priorities rather than individual needs or preferences. This top-down implementation strategy often overlooks the importance of user buy-in and proactive exploration.
The pervasive narrative surrounding AI frequently amplifies fears and anxieties about job displacement. This constant barrage of messages about AI potentially replacing human roles creates an environment of apprehension rather than excitement. Consequently, the perception of AI is often one of resistance to change and deep-seated concern about one’s place in a rapidly evolving world. This is compounded by the feeling that these changes are occurring without their direct input or control.
Recent studies have begun to quantify the impact of AI integration on work. A comprehensive analysis indicated that while AI might appear to boost productivity in certain areas, it can also lead to unexpected increases in workload and the need for constant vigilance. For instance, email and chat communication times have seen significant upticks, as has the use of various business tools. Alarmingly, some data suggests an increase in working hours, including weekends, and a reduction in focused work time. Moreover, the study pointed to a rise in costly mistakes and the time spent rectifying AI-generated errors. This suggests that rather than reducing the overall burden of work, AI can, in many instances, intensify it.
At best, AI features may be met with passive acceptance or a resigned nod. At worst, they can evoke significant concerns, doubts, and a healthy dose of skepticism. This caution is often rooted in the inherent unpredictability and unreliability of AI, which distinguishes it from other, more consistently performing software features. Unlike traditional software functionalities that users can rely on to perform as expected, AI’s probabilistic nature can introduce an element of uncertainty, making it perceived as a potential liability.
The vision of a future filled with AI-generated art museums, AI-powered refrigerators, or AI-assisted hotel check-ins does not resonate with the desires of most individuals. Similarly, the prospect of romantic AI partners or the active management of a swarm of AI agents operating within personal financial accounts is met with apprehension rather than enthusiasm. The idea of constantly interacting with a disembodied "magical box" also fails to capture the human desire for meaningful interaction and a sense of control.
The AI People Truly Need: Reliability, Augmentation, and Relief from Tedium
The common comparison of AI’s unreliability to human fallibility misses a crucial point: users do not compare AI to people; they compare AI features to other features. When one AI-driven function within a product proves unreliable while a similar function in a competing product performs flawlessly, users will invariably choose the latter. The critical factor is not whether a feature is powered by AI, but whether it is consistently reliable and effective.
Discussions surrounding AI often emphasize the speed of delivery. However, for many, the value of increased speed is diminished if it comes at the expense of quality or thoughtful execution. Users desire to perform tasks well, with sufficient time for reflection and sound decision-making. They also seek to derive satisfaction and enjoyment from their work, rather than simply accelerating its completion. The current trajectory of technological implementation, characterized by rapid, often superficial changes, risks eroding this sense of reward and achievement.
Human needs and desires remain remarkably consistent. After years of technological evolution, individuals continue to prioritize features that are fast, accessible, reliable, and predictable. Crucially, they seek tools that augment their existing capabilities rather than seeking to entirely replace their established workflows. The ideal AI, in this context, is one that takes over the most mundane, tedious, and unrewarding aspects of a job, freeing up human cognitive resources for more engaging and creative endeavors.
The potential impact of AI on the job market is a significant concern, with numerous roles identified as being exposed to automation. However, many of these professions contain a core component of rewarding, unique, and creative work that relies on human taste, perspective, and intuition. When AI can effectively automate the repetitive and less cognitively demanding aspects of these jobs, it presents a clear advantage for both individuals and organizations. This automation of drudgery can significantly enhance productivity and contribute to a more positive daily work experience.
The value proposition of AI becomes significantly clearer when it automates tasks that are tedious and mentally draining. For AI to achieve this effectively, it must be seamlessly integrated into existing workflows, rather than feeling like an add-on. Furthermore, these AI functionalities must align with the established mental models that users have developed over years, or even decades, of experience. AI should adapt to human cognitive processes and decision-making frameworks, rather than forcing users to fundamentally alter their approach.
The branding of these features—whether as "AI," "smart," or "automation"—is secondary to their performance. The critical requirement is that they function effectively for the end-users. This necessitates clear communication regarding the use cases where AI genuinely provides assistance, and fosters an environment where users are encouraged to discover and leverage these benefits proactively.
Paradoxically, the most effective tools are often not "AI-first" but rather "AI-second." These are the subtle, humble, and ambient technologies that operate supportively in the background, taking over tasks that are otherwise dull and unnecessary. They enhance the human experience by removing friction, rather than demanding constant attention or a significant shift in user behavior.
Bo Young Lee, in a widely shared sentiment, articulated this desire powerfully: "I don’t want to read books written by AI. I don’t want to gaze upon paintings by AI. I don’t want AI to teach my children. I don’t want to have an AI therapist. I don’t want AI making my medical decisions. I want AI to do all the physical and mental labor that taxes me so I can read books written by humans and go to art galleries to engage with art made by humans. I want AI that makes my life easier rather than forces me to change myself." This perspective encapsulates the aspiration for AI to serve as a facilitator of human experience, not a replacement for it.
Conclusion: Towards an "AI-Second" Future
While some may view this perspective as old-fashioned, the underlying sentiment reflects a deep appreciation for human connection and interaction. The value of human stories, thoughts, emotions, and enthusiasm remains unparalleled. While AI can be incredibly helpful in specific contexts, the preference for spending time with a human, despite their imperfections, is a testament to the enduring importance of human relationships.
Ultimately, people do not require more AI in their lives as a direct replacement for human interaction or as a constant presence. Instead, they need AI to effectively automate the monotonous and time-consuming aspects of their daily routines. This automation would free up valuable time and mental bandwidth, allowing individuals to engage more fully in activities they genuinely enjoy and find fulfilling. This shift would not necessitate increased interaction with AI itself, but rather would facilitate more meaningful and abundant time spent with other people.
The future of AI integration lies not in a relentless push for more AI features, but in a thoughtful and user-centric approach that prioritizes reliability, augmentation, and the alleviation of tedium. By focusing on "AI-second" solutions that seamlessly support human endeavors, businesses can move beyond the illusion of demand and deliver genuine value to their users.
This article is brought to you by Design Patterns For AI Interfaces, offering friendly video courses on UX and design patterns by Vitaly, with a live UX training session scheduled soon. A free preview is available.
