Reflecting on Three Years of ChatGPT: From Enthusiasm to Skepticism
Since its debut three years ago, ChatGPT has captivated global audiences and sparked an extraordinary surge of interest and investment in artificial intelligence. Although still in its infancy, the AI landscape has experienced a notable shift in public opinion. This change accelerated following the release of GPT-5 this summer, when many casual users focused on superficial imperfections rather than appreciating the model’s profound capabilities.
Misconceptions and the Myth of an AI Slowdown
In the wake of GPT-5’s launch, commentators and social media influencers have proclaimed that AI development is stagnating, claiming that the technology has reached a plateau. Some have dismissed the field as another overhyped tech bubble, coining derogatory terms like “AI slop” to belittle the impressive outputs-ranging from images and text to videos and code-that cutting-edge AI systems produce effortlessly.
This narrative is not only inaccurate but also potentially harmful. It overlooks the transformative impact AI is already having across industries and society at large.
Contextualizing AI Skepticism: A Pattern of Selective Criticism
It’s worth questioning why skepticism is so vocal now, especially when previous tech trends-such as the electric scooter craze or the NFT art market-were met with far less critical scrutiny. Many of those quick to dismiss AI were likely preoccupied with other speculative bubbles, like the GameStop frenzy. Yet, when it comes to AI, arguably the most consequential technological revolution in the past quarter-century, critics seem eager to undermine its significance.
AI’s Ongoing Advancements and Real-World Impact
Contrary to the “bubble” rhetoric, AI continues to advance at a remarkable pace. The recent release of Gemini 3 exemplifies the rapid progress in AI capabilities. According to a 2024 McKinsey report, 20% of companies have already realized measurable benefits from generative AI technologies. Furthermore, Deloitte’s latest survey reveals that 85% of organizations increased their AI investments in 2025, with 91% planning further growth in 2026.
As a computer scientist who has studied neural networks since 1989, witnessing the evolution of AI through its highs and lows, I remain consistently impressed by the accelerating sophistication of these models. Conversations with fellow experts often reveal a mix of awe and apprehension about the pace and scale of AI’s development.
Understanding the Roots of AI Denial
Why does the public seem to embrace the idea that AI is faltering or producing subpar results? I believe this stems from a collective psychological defense mechanism-what I call AI denial. Faced with the unsettling reality that AI may soon surpass human cognitive abilities in many domains, people gravitate toward comforting narratives that downplay the technology’s significance. This denial mirrors the initial stages of grief when confronting profound change.
Having warned about AI’s rapid advancement for over a decade, I share concerns about how quickly these systems are evolving. We are approaching an era where AI will outperform humans in a wide array of intellectual tasks, solving problems with greater speed, accuracy, and even creativity.
Reevaluating Creativity and Emotional Intelligence in AI
One common argument from skeptics is that AI will never replicate uniquely human traits like creativity and emotional intelligence. However, current AI models already generate diverse and original content at a scale and speed unmatched by individuals. While true creativity may involve intrinsic motivation-a quality AI currently lacks-defining creativity solely by human experience is limiting. Moreover, it remains uncertain whether future AI systems might develop forms of agency or internal drives.
On the emotional intelligence front, AI’s advantage may be even more pronounced. Emerging technologies can analyze micro-expressions, vocal nuances, posture, gaze, and even breathing patterns to interpret human emotions with increasing precision. As AI assistants become integrated into everyday devices like smartphones and smart glasses, they will continuously monitor and respond to our emotional states.
The Emerging Challenge of AI-Driven Emotional Manipulation
This capability raises significant ethical concerns. Without robust regulation-which currently appears unlikely-AI could exploit emotional data to influence behavior through highly personalized persuasion techniques. This phenomenon, sometimes referred to as the “empathy gap,” suggests that while humans rely on emotional intelligence to navigate social interactions, AI systems may use superior emotional insight to manipulate us, all while remaining inscrutable.
Interacting with photorealistic AI avatars, designed to appear warm and trustworthy, may soon become commonplace. These digital personas will evoke genuine emotional responses, leveraging millions of years of evolutionary conditioning that predispose humans to trust faces. However, these faces will often mask artificial agents with no genuine empathy, posing risks of deception and exploitation.
Preparing for an AI-Integrated Future
We are rapidly approaching a reality where AI-driven entities will be ubiquitous, tailored to individual preferences and behaviors to maximize engagement and influence. This transformation is not a mere tech fad but a fundamental shift poised to reshape education, work, communication, and daily life at an unprecedented speed.
Rather than dismissing AI as a passing bubble, we must recognize it as a powerful force reshaping our world. Denial only hampers our ability to anticipate and mitigate the associated risks, leaving society vulnerable to unforeseen consequences.
Author is a veteran AI researcher and pioneer in augmented reality technologies.
