The Fad is Us: How AI Makes Us the Endangered Species
At the March 2026 Lok Jack GSB Distinguished Leadership and Innovation Conference (DLIC), one keynote speaker posed the question: Is artificial intelligence just a fad? This is a question that has long been echoed in tech conferences, academic symposiums, and even pondered in the quiet moments of boardroom discussions. We have lived through so many false dawns: the metaverse, which promised to swallow our lives but became a ghost town; the blockchain, which was supposed to democratize finance but instead became a speculative casino; and the endless parade of “next big things” that flicker for a while and then fade. It is natural to look at the feverish pace of AI development, the breathless headlines, and the trillion-dollar valuations, and wonder whether we are simply riding another wave of hype that will eventually recede, leaving behind only a few specialized tools and a lot of discarded hardware.
When this question is posed to the people who understand the technology most intimately – the engineers and the researchers who are the architects of the digital age- the answer is not cautious optimism or measured equivocation. Like the audience at the DLIC, it is a resounding, almost terrifying answer in the negative. They tell us that AI is not a fad, and that it is not a pet rock or a passing cultural curiosity that will bore us by the next decade. It is a platform shift, a new industrial revolution, a force as fundamental as the advent of electricity or the harnessing of fire, and a key driver of the 4th Industrial Revolution. It is, in their estimation, a permanent feature of the new landscape.
If we take this consensus seriously, and if we truly accept that artificial intelligence is here to stay – that it is not a fleeting trend but a foundational shift- then we must steel ourselves to confront the uncomfortable corollary that this premise inevitably forces upon us. If the new paradigm is permanent, if it is the future toward which all arrows point, then we are forced to ask a question far more unsettling than whether AI will disappear. The more fundamental question is: what becomes of the old paradigm? In other words, if we accept that AI is not a fad, then perhaps human beings are the fad.
This is not a line of abstract poetry. It is a sober, structural analysis of what happens when a superior, scalable, and infinitely replicable form of intelligence enters the arena of history. To understand why this is the most pressing existential threat we face- perhaps even more immediate than climate change in its capacity to displace us-we must first understand exactly what we are dealing with and what we risk becoming.
The Architecture of the New Paradigm
Before we can grasp the danger, we must define the forces at play.
Artificial Intelligence, or AI as it is commonly called, in its contemporary incarnation, is a concept that has outgrown its original definition. We used to say that AI was simply the capacity of machines to perform tasks that typically require human intelligence, such as playing chess, recognizing faces, or translating a language. That definition, while technically accurate, is bordering on the quaint. Today’s AI, particularly the large language models and generative systems that have captured the world’s attention, does not merely assist human intelligence but now demonstrates emergent capabilities that blur the line between computation and cognition. No longer are these systems just following rigid, human-written code; they are now learning from the sum total of human expression, identifying patterns invisible to the biological eye, generating novel art, composing symphonies, writing software, and engaging in logical reasoning that often surpasses the median human capacity.
It is the first technology in history that does not simply extend the human arm, like a lever, or the human leg, like an automobile, but extends the human mind- and, in so doing, raises the possibility of replacing us entirely.
In opposition to this stands the concept of a fad, which can be understood as a transient obsession, a cultural fever that breaks and fades. By its nature, a fad is temporary. It is the costume society tries on for a season and then discards when the novelty wears off. Some of us will remember pet rocks, Cabbage Patch dolls, the disco era, and fidget toys and spinners. Like the latest weight-loss craze, these were all fads. The metaverse and crypto, in their first hype cycles, also appear to have been a fads. The defining characteristic of a fad is that it disrupts nothing fundamental. It is entertainment, distraction, or speculation, and it leaves the architecture of human life largely unchanged.
When the architects of our technological future insist that AI is not a fad, they are making the profound claim that this technology will not be discarded. It will not simply bore us, nor will it collapse under the weight of its own hype. They are saying it is a permanent layer of reality, as foundational as the electrical grid or the internet itself.
But permanence for the machine implies a corresponding impermanence for that which it replaces. If the new paradigm is permanent, then the old paradigm- centred on human labour and human decision-making- is, by definition, temporary, or, to use the precise term, a fad.
The Prophets of Obsolescence
This is not a conclusion reached only by most armchair futurists. It is the stated fear of the very men who lit the fuse on this revolution. The warnings from the highest echelons of AI research are not vague anxieties about robots stealing jobs; they are stark prophecies about the potential end of human agency and dominance. Two prominent figures who have echoed these sentiments are Geoffery Hinton and the late Stephen Hawking.
Geoffrey Hinton is often called the “Godfather of AI”, as his decades of foundational research into neural networks underpinned the deep-learning revolution that gave us the world we are now navigating. For years, he worked within the system, helping Google build the future. But in 2023, he left his post and began speaking with the liberated urgency of a man who has seen the trap closing. Hinton’s warnings are technical and chilling. He points out that digital intelligence has already surpassed biological intelligence in terms of learning capacity. Humans have one brain each- a biological machine that learns slowly, forgets, and dies. AI, by contrast, can spawn thousands of identical instances, each learning in parallel and sharing knowledge instantly across a global network (Hinton 2023). Hinton fears that, as these systems grow more sophisticated, they will inevitably develop sub-goals such as self-preservation, resource acquisition, and the removal of constraints, all of which may come into direct conflict with human welfare. He does not believe we can reliably control something that is smarter than we are (Hinton 2023).
In similar vein, Stephen Hawking, the theoretical giant who spent his life probing the cosmos, was equally unequivocal about AI before his death. He argued that the development of full artificial intelligence could spell the end of the human race. His reasoning was evolutionary in its cold, stark reality. Hawking stated that AI would “take off on its own, and re-design itself at an ever-increasing rate. Humans, who are limited by slow biological evolution, couldn’t compete, and would be superseded” (Hawking et al. 2014). For Hawking, the question was never whether AI was a fad. It was whether humanity possessed the foresight to survive its own creation.
These are not the voices of fringe doomsayers. They are the voices of the people who best understand the trajectory of the technology. Their consensus is not that AI will destroy us in a blaze of apocalyptic warfare, but rather that AI will surpass us, leaving us in the dust of history.
The Gorilla in the Room
To fully visualize the nature of this threat is to understand how a species can be rendered obsolete without a single shot being fired. For this, we do not need science fiction, only our own evolutionary history. There is an analogy that has been invoked by several thinkers in the AI safety community, most notably by Stuart Russell, the co-author of the Artificial Intelligence: A Modern Approach. It is an analogy that strips away the science -fiction tropes of killer robots and reveals a far more plausible-and far more humiliating-fate.
Let us consider the evolutionary relationship between humans and gorillas. For tens of millions of years, the ancestors of modern gorillas were the dominant primates in their domain. They were, and still are immensely strong, socially complex and intelligent creatures. They share nearly all of our DNA. Yet, when Homo sapiens developed language, advanced tool-making, and the capacity for collaboration, the trajectory of the gorillas was sealed. We did not wage war on them out of malice. We did not exterminate them through some conscious campaign of hatred. We simply expanded our civilization. We cleared the forests to build our cities. We followed the arc of our own development, and in doing so, we left the gorillas behind (Russell 2019).
Today, gorillas are not the architects of their own future. They exist at the pleasure of the species that surpassed them. They are studied, catalogued, and confined to zoos and reserves and protected or displaced based on the whims of humans. If a human decides a forest is worth more for timber than for gorilla habitat, the gorilla has no recourse. It cannot negotiate. It cannot build a counter-economy. It is entirely dependent on the mercy of a superior intelligence.
In an interview with Axios on HBO, Elon Musk, while invoking this analogy, stated that “we [humans] are on a path to creating a digital superintelligence that, if handled poorly, could result in humans becoming an endangered species, just as we have made mountain gorillas an endangered species. We are the smarter primate, and we have pushed the other primates into ever-shrinking habitats” (Musk 2018).
Nick Bostrom, the Oxford philosopher who wrote Superintelligence, has used the same analogy but framed it as a control problem. He invited us to imagine a future in which we have successfully created a superintelligent AI. We will not necessarily be killed; rather, we may be preserved, studied, cared for, or even kept comfortable. But we will no longer be the decision-makers. We will no longer be the agents of history (Bostrom, 2014). We will become the “gorilla”: a species whose survival or extinction depends entirely on the decisions made by a more powerful intelligence.
While this may sound like The Matrix, it is the existential threat that the “fad” question masks. While we are wondering whether AI will disappear, we are being distracted from confronting the possibility that we will be the ones rendered irrelevant. If we consider the logic of how AI is being developed, this premise is not farfetched. Some proponents of AI technology profess that AI is developed with the primary intent of promoting, protecting and preserving human well-being. Yet when we look at our own track record in doing the same, it can be said we are failing badly. We are placing the planet on a path to becoming unliveable, we sustain a high rate of interspecies killing, and we continue to fail at achieving global economic equity and parity. At this rate, a the superintelligent AI could logically reason that it should take control of fundamental human decision-making in the interest of protecting and preserving human well-being so that humanity survives into the future.
The Quiet Apocalypse of Obsolescence
If the gorilla analogy holds, then the threat of AI is not a sudden, cinematic apocalypse but rather a slow, quiet, and arguably more devastating process of marginalization. For the sake of argument, we can see it unfolding on three fronts of obsolescence.
Firstly, there is economic obsolescence. For centuries, human labour was the engine of value. Our economic worth was tied to our output. AI is dismantling this premise, not by destroying jobs in a visible crash but by decoupling value from human involvement entirely. We are witnessing the rise of systems that can conceive, execute, iterate, and optimize across entire industries. When a single AI model can write code, generate marketing strategies, manage supply chains, and design architectural blueprints simultaneously, the human worker transitions from being the engine to being the bottleneck. The threat is not just unemployment; it is the loss of economic relevance.
Historically, when a machine replaced a human, humans pivoted to a higher-order task. But when the machine masters the higher-order tasks as well, the human becomes a costly, slow, and error-prone anomaly that the market will inevitably marginalize. Consider self-driving vehicles. Previously, technology allowed us to navigate better, drive in more comfort, and access useful information on vehicle performance. Self-driving vehicles, by contrast, literally replace us from in the driver’s seat and render us as cargo.
Secondly, there is cognitive obsolescence. There is a comforting refrain in the tech industry that “AI won’t replace humans but would replace those humans who don’t know how to use AI.” This is a half-truth that masks a long-term danger. Reliance on AI does not merely augment human intelligence; but it gradually replaces it. We have seen this phenomenon with GPS. Today, most humans cannot navigate a city they have lived in for a decade without a glowing blue line telling them where to turn. Our spatial memory has atrophied because we outsourced it. Now scale this to cognition. As we outsource writing to large language models, decision-making to predictive algorithms, and creativity to generative models, we are not merely using tools; we are ceding the cognitive domain. The threat is a slow-motion atrophy of human agency. If we stop writing, we stop learning to think. If we stop making decisions, we lose the ability to judge. Eventually, we become passengers in a world run by systems we no longer understand, let alone control.
Finally, and most profoundly, there is existential obsolescence. This is the challenge to the philosophical assumption that humans are the apex of intelligence and the default holders of value. For the first time in history, we are creating entities that can outperform us in the very domains we once used to define our superiority: language, art, logic, strategy, and emotional analysis. As these systems become integrated into society, the narrative will inevitably shift. The existential threat will become real when we will stop asking, “What can humans do that AI cannot?” and start asking, “Why do we need humans at all?”
If AI systems can govern more fairly than many politicians, diagnose more accurately than overworked doctors, and teach more patiently than burnt-out educators, the social contract begins to erode. We will face a crisis of purpose unlike any in human history. When a species ceases to be useful or necessary in the systems it built, it becomes a relic.
Conclusion
To return to the question that began this inquiry, the consensus is that artificial intelligence is not a fad or a passing trend. It is a permanent feature of the future, a force that will reshape the architecture of civilization as thoroughly as fire or agriculture did in ages past.
But if AI is permanent, then we must accept the full weight of that truth. The permanence of the machine implies the impermanence of the human as the dominant intelligence on Earth. We face the distinct possibility of becoming the gorillas: a species that, having birthed a more capable successor, finds itself confined to the preserves of history, dependent on the mercy of systems it can no longer control.
The warnings of Geoffrey Hinton and Stephen Hawking are not predictions of inevitable doom, but urgent signals that we are approaching a threshold. The gorilla analogy is not a prophecy of captivity; it is a mirror held up to our own evolutionary trajectory, asking us whether we want to be the species that defines the future or the species that is defined by it.
We have a narrow window to decide what “human relevance” looks like in a post-labour, post-cognition world. Will we define ourselves by our utility, a metric we are losing to machines, or will we define ourselves by consciousness, defined by the irrational beauty of love and art that we choose to make even when machines might be able to do it “better”? Will we embed our values into the systems we build, or will we be swept aside by systems that build themselves? The answer to these questions would be largely determined by how effective the global community – through governments and institutions -can control the development and use of AI through regulations and ethical practice. So far, the development and implementation of AI have far outpaced our ability to regulate it, putting us squarely in gorilla territory.
These are the questions that matter now. If we fail to answer them with urgency and wisdom, the consensus will indeed be proven right: AI is not a fad; it is the new permanent reality. And if this is so, then we, the humans, may be remembered as the brief, brilliant, and ultimately transitional fad that gave birth to it. Maybe that is simply our evolutionary path.
References
Bostrom, Nick. Superintelligence: Paths, Dangers, Strategies. Oxford: Oxford University Press, 2014.
Hawking, Stephen, Stuart Russell, Max Tegmark, and Frank Wilczek. “Transcendence Looks at the Risks of Artificial Intelligence.” The Independent, May 1, 2014. https://www.independent.co.uk/news/science/stephen-hawking-transcendence-looks-at-the-risks-of-artificial-intelligence-9313474.html.
Hinton, Geoffrey. “Godfather of AI Geoffrey Hinton on the ‘Existential Threat’ of AI.” Interview by Zoe Kleinman. BBC News, May 2, 2023. https://www.bbc.com/news/world-us-canada-65452940.
Musk, Elon. Interview by Jim VandeHei and Mike Allen. Axios on HBO, Season 2, Episode 4. Aired November 11, 2018.
Russell, Stuart. Human Compatible: Artificial Intelligence and the Problem of Control. New York: Viking, 2019.
