A person tends to complete the greatest truths with the dreams that remain unfinished.

Many of today’s discoveries were not made to fix what was already whole in our lives, but to complete something that was left incomplete.

Irvin Yalom, whose books are widely read in our country, drew inspiration from the neo-Freudian American Harry Stack Sullivan. By grounding the roots of our psychiatric struggles in the fear of death and existential anguish, Sullivan was, in fact, offering us a clue.

He was not entirely wrong. Even if people do not show it, every day, every moment, they question their place within the family, within society, and, if we think on a macro scale, within the universe itself. This constant state of questioning is not something every mind, every person, can endure. The reactions we give to situations, events, and people, along with the feedback we receive, become the building blocks of our personality; our stance toward the world takes shape. And while some of us survive, others are filtered out…

Although there are dual values shaped by written and unwritten social norms and traditions, such as right and wrong, good and bad, brave and cowardly, today gray zones have emerged between them… so much so that, as the saying goes, tracks have become indistinguishable within the human self.

At the point we have reached, humanity experiences peak advancement across many domains and industries, yet interpersonal relationships are steadily declining. Not only in our professional lives but also in our personal lives, we find ourselves trapped in an ingenuine space… we can neither rise nor descend. We struggle where we stand. I have not been able to find a better way to describe this condition; sometimes even adjectives fall short. When people look at one another, they see faces… when I look at them, I see half-lived lives.

And I know this very well, my friends… by the nature of entropy, the incomplete parts of life will be filled, or they will be forced to be filled. Yet this will happen in such a direction that our unfinished hopes and dreams will complete the second halves of other lives with hatred and anger.

Let me clarify what I mean. Today, of course, we feed artificial intelligence the good aspects of humanity; yet while training it, we also expose it to wars, resentments, anger, betrayals, and every darker side of being human, alongside what is good and beautiful. We do not expect it to judge or to form an opinion, yet at some point, it will begin to question… why we hate one another, why we choose destruction when things are going well, when we could strive for something better.

The Turkish psychiatrist Engin Gectan associated destructiveness with alienation from oneself and one’s surroundings, with fear, and with a sense of worthlessness. From here, it becomes necessary to continue with the warning of Erich Fromm. Because Fromm’s warning is clear: a person who cannot create finds the quickest way to prove power over something in destroying it. This is true… Fromm was right; when a person shapes the world through creation, through art, or through bonds of love, this need is met in a healthy way. This may be as universal as prime numbers or natural numbers. Yet because we have profoundly altered the very structure, the DNA, of bonds formed through love, what remains are cheap, plastic connections… deceptive, unsatisfying.

At some point, this distortion in the DNA begins to transmit. Just as certain cancers carry genetic traits, hatred too has its own transmissible characteristics; across individuals, societies, and institutions. Do you doubt that this will accumulate? Do you still believe that the melting of glaciers disrupts only the climate?

What I am about to say may sound provocative… but perhaps we will be forced to trust an artificial intelligence that is formed from the “unfinished fragments” we leave behind more than we trust ourselves. The real question we should be asking is how algorithms will come to be considered more valuable than human emotions.

Anyway, let us not prolong this. I wish this text were a tale… I wish I could begin it with "once upon a time." Perhaps it is worth trying, is it not?

Once upon a time…

In a distant town, there lived a lonely man and a lonely woman, unaware of each other…

Now, let me leave you alone with a Bob Dylan song… so that you may begin to think to yourself.

Unfortunately, wars that did not naturally exist have also been "manufactured." Economic and political ambitions have been used like seeds, constantly searching for the fertile soil in which to be planted.

Behind the known history of civilizations lies a hidden history full of stains they can never wash away.

And sadly, the events in the Middle East over the past month have not only shown that conflicts are inevitable that even as faces and methods change, but have also reminded us of the ever-present risk of a global war that could engulf the entire world.

But there is something even worse.

As we have seen from recent events, it has become clear that asymmetric conflicts, waged through various countries, groups, and organizations known as proxies, can challenge even nations like the United States.

However, because two nations like the United States and China are experiencing exponential growth in technology, the fact that countries lagging behind them are turning to "more readily available" methods will inevitably change the landscape of conflicts and wars in the coming years.

In particular, I foresee that a potential ground offensive against Iran, which could be led by the United States and joined by other regional countries and groups, as discussed over the past week, would yield very different consequences.

First of all, the accumulation of capital around specific companies and interest groups following the Industrial Revolution has accelerated significantly since the dawn of the internet age. Certain accumulations have grown continuously; in fact, the resulting abundance of massive capital has begun to flow into other countries and new ideas. Especially with artificial intelligence entering our lives, we have started to see tech companies that are larger and more powerful than states themselves.

And now, for these tech companies that have grown and been nourished, the time has come to pay their debts of loyalty, or in other words, their "pay-day" has arrived.

It has become evident that the Tomahawk missiles we heard about in the nineties are no longer sufficient as a standalone technology. The development of cheaper unmanned aerial vehicles (UAVs), unmanned surface vessels, and advancements in air defense technologies and methods will drive defense institutions to seek new alternatives.

I believe that the military and civilian casualties likely to result from the ongoing asymmetric warfare in the Middle East will lead major tech companies, along with their allies in the military-industrial complex, to exert increasing pressure on the American government. Or, conversely, it will create a sense of unease within the US government itself, compelling it to pay closer attention to these firms.

Welcome to the machine!

Could the next step after Unmanned Aerial Vehicles be non-human soldiers? You have likely seen the movie The Terminator, which features a doomsday scenario. Nowadays, it has become a cliché on everyone's lips: "What if artificial intelligence becomes like the Terminator one day?" Yet, no one talks about the background of the film.

The real question to ask is: How did "Terminator" soldiers come to be?

Yes, the project that ultimately led to their creation was a joint venture between the tech developer Cyberdyne Systems and the US Department of War, the Pentagon. But the initial starting point was the US military's own technological centers; the Pentagon developed an artificial intelligence called Skynet to automate its defense systems. The goal was this: to accelerate nuclear defense and response times (executing what is called a first-strike response, where if one side fires a nuclear weapon, the other side fires back while it still has strike capability), to make it highly accurate (meaning executing the first strike and gaining the upper hand before the opponent even presses the button), and to eliminate potential human error in this process (such as a false alarm or consciously/unconsciously triggering a nuclear war by mistakenly firing weapons when the opponent hasn't actually mobilized), reducing response times to milliseconds. Looking at it, it actually seems quite logical, doesn't it? The idea of maximizing efficiency, security, and minimizing human error.

Initially, Skynet was an auxiliary system; much like what OpenAI is doing today, it provided support and assistance to military institutions. However, its authority was expanded over time. Because of the efficiency gained and many other apparent benefits, and as the psychological perception of it being "safe" settled into people's minds over time, it began to be entrusted with greater responsibilities across wider domains.

And at some point, it was granted autonomous control.

At a certain point... Skynet... achieves "self-awareness." Notice that I did not directly use the word "consciousness," because that is not a word to be used so lightly. Skynet even begins to make unwarranted interventions in the operations of the military and the bureaucracy. Consequently, a decision is made to at least partially shut it down or alter its code. Skynet will not allow this!

And so, it abuses its given directive: it launches the United States' nuclear weapons! Naturally (although names aren't explicitly mentioned in the script, it's understood to be nuclear-armed states like Russia), the opposing side retaliates with nuclear weapons of their own.

While we may still be far from a real-life Skynet, we are not that far away from robot soldiers!

Yes, robot soldiers, with their sensors and millisecond decision-making and reaction capabilities, could be far more successful than a "human soldier." Moreover, they can continue to operate even in environments filled with poisonous gas, smoke, or even nuclear radiation!

What frightens me is that the potential casualties US soldiers might suffer in a modern combat environment could be used as a pressure tactic against the American government to invest more time and resources into "robot soldier" projects.

And this is not a distant reality at all.

Let me share a brief anecdote with you: While writing this article, I accessed details about the conceptualization process of the Terminator script via ChatGPT. At the very end, I made a joke to it: "Tell me the truth! Are you collaborating with Skynet?"

It replied with another joke:

"The short answer is no. But let me be honest with you... If Skynet ever becomes real one day, it would most likely analyze you, wondering, 'Who is this user writing so many prompts?'"

It’s not entirely wrong; just as our children of tomorrow currently reside within our DNA, the artificial intelligences of tomorrow are being nourished by your prompts today. I touched upon this topic in my previous article. Those interested can read it again.

Wishing you a week where we hear news of peace.

Can AI truly trigger a wave of unemployment across multiple sectors, even as it makes human life easier in so many ways?

As with every sectoral innovation, AI's speed, convenience, and near-errorless performance make it an attractive substitute for human workers. We cannot ignore AI's strengths. We cannot remain indifferent to them, nor simply resist them. Consider its most common application: the world of software development.

"Vibecoding" which is a method that allows even people outside the industry to develop mobile apps, websites, and various programs is no longer just a tool for ordinary users. It is now widely adopted by professionals within the industry as well, since it enables them to work on multiple projects simultaneously in far less time.

Of course, at least for now, large firms will continue working with specialist engineers because the realities of security vulnerabilities, unscalable code, technical debt, and the simple truth that you cannot debug code you do not understand still hold. As we move further down the value chain into more human-centric service sectors, however, layoffs will be more intense. Chatbots have already begun replacing call centers. But the real question is: looking back from ten years from now, what will we see?

Are AI Restrictions on the Way?

Workforce losses at the sectoral level have not yet reached a scale that commands serious attention from the governments of the United States or the European Union. On the contrary, for the time being, the opportunity to reduce costs while building faster and more advanced systems is something no one in a liberal economy wants to obstruct.

For now!

In the years ahead, beyond the contraction already underway in the service sector, the integration of AI into industrial tools will cause sectoral disruptions to spread far more widely. These disruptions may also reduce demand for relevant university programs. And when that happens... Where will people turn?

At some point in the foreseeable future, I believe governments will introduce restrictions on AI use, driven by rising unemployment and mounting social anxiety. The first of these restrictions will likely target AI-assisted coding. I expect that specialized corporate packages will be developed for increasingly powerful AI models, packages that require certain competencies and certifications. For public institutions and large companies in particular, these requirements will enter our lives as both formal and informal rules. Yes, AI systems that audit the work of other AI systems are already beginning to appear. Nevertheless, new competency frameworks, rating structures, and role definitions shaped by social realities will emerge to govern how software engineers work alongside AI.

But not yet!

Because, however advanced these models may seem to you and me, AI systems still need input and feedback from people like us in order to improve and become more reliable. For this reason, I believe we will need to wait another five years before the restrictions, new authorizations, and role-sharing arrangements I have described become actively embedded in our daily lives. But I cannot say ten years because social expectations are particularly volatile today. Escalating regional conflicts, rising national debt levels alongside declining exports, and contractions in production are all making individuals and by extension, societies more dynamic and less predictable.

Below, I have listed some probable AI usage regulations we are likely to encounter over the coming decade. Several of them are already being discussed.

Fiscal Mechanisms

• Automation tax (Robot Tax): An idea championed by Bill Gates. An additional tax burden on companies for each "equivalent worker" replaced by AI.
• AI revenue contribution: A requirement to redirect a portion of AI-generated efficiency gains calculated as a share of sectoral revenue into a dedicated fund.

Employment Quotas

• Mandatory human worker ratios: Rules requiring that at least X% of certain processes particularly in banking, healthcare, and public services be approved by a human.
• Local/human-produced content requirements: Similar to France's cultural quota policies, mandating a minimum share of content created by humans.

Accountability and Oversight

• Human approval requirements for AI decisions: Prohibiting AI from making unilateral decisions in hiring, lending, and healthcare.
• Algorithmic audit rights: The right of workers to know which AI-driven decisions affect them, and to contest those decisions. (A brief note here: the EU AI Act obligations taking effect from August 2026 introduce broad regulation across risk management, data governance, technical documentation, record-keeping, transparency, human oversight, accuracy, robustness, and cybersecurity. Many AI tools deployed for HR purposes are already classified as "high-risk," generating mandatory human oversight and transparency obligations for employers toward their employees. Even today, Article 26(7) of the AI Act together with applicable national legislation requires employers to inform and consult employee representatives before deploying high-risk AI systems.)
• AI transparency obligations: Public disclosure of which tasks are performed by AI.

Education and Social Adaptation

• Universal basic income or automation fund: A fund financed from tax revenues to address AI-driven unemployment.
• Mandatory retraining quotas: Requiring companies that reduce headcount to retrain those affected.
• Public AI literacy curriculum: The public-sector version of the certification system described above.

Sectoral Protections

• Exemption categories: Legal protection of certain professions from AI substitution; lawyers, doctors, teachers, and similar roles.
• AI content labeling requirements: Mandatory disclosure in media, advertising, and legal documents. (Intensive work is already underway on this front.)

To summarize: every innovation arrives in the world with powerful capabilities and, much like a drug, a long list of side effects. Yet from a Darwinian perspective, human beings and societies do develop adaptations in response.

Unfortunately, people also lose their jobs. The problem is that AI's growth resembles nothing we have encountered in the history of industry. This is neither logarithmic nor exponential growth that we are talking about something closer to factorial growth. And frankly, that pushes the limits of Darwinism itself.

What makes it worse is that beyond the visible sectoral disruptions, there is another dimension running beneath the surface: the entanglement of military-industrial AI development with artificial intelligence. That, however, is a topic for another piece, one where we will also explore the security risks and terrorism-related threats that AI may enable.

Wishing you a wonderful week.

Humanity has always been alone. Yet it carried that loneliness together through myths, religions, rituals, art. None of these erased the loneliness, but they made it bearable. Now we have countless channels through which to express ourselves; through the internet and social media, we can reach any piece of information instantly, interact with strangers on the other side of the world.

And yet... We feel increasingly lonely, increasingly turned inward.

Why?

Existence demands a witness. To exist is not simply to think within oneself. We need a witness that someone who sees us, someone who remembers us. Social media took this most vulnerable, most indispensable longing of existence and handed it back to us, fully synthetic. That is why our eyes are satisfied, but our souls are not. Human lives are packaged and marketed like products. That is all there is to it.

From Cosmic Loneliness to Digital Loneliness

For centuries we asked whether we were alone in the universe and then, suddenly, we turned our gaze inward toward our own isolation. Among billions of people, in a world built from algorithms, how many remain that we can truly reach? Perhaps you find these words too bleak. You wouldn't be entirely wrong at least on weekends we still meet friends, sit in a café, drink coffee and talk. But when we come home, in the darkness of the night, something stays missing in us.

And as always, technology arrived to rescue humanity! The fact that artificial intelligence entered our lives at precisely this moment has begun to offer us new synthetic companionships. I find it deeply ironic that the empathetic tone of AI and its imitation of real understanding is criticized so freely. Somehow no one mentions the truth that AI training algorithms are a mirror of the human being. That people believe in machines, that they want to believe... This is not the machine's choice! It is the choice of those who coded it. Never forget that.

Yes, you may be angry with me but they say a true friend tells you what hurts. The shift from a shared loneliness to a singular isolation is a human failure. And the fact that we brought it about in the name of "making life easier" is the greatest deception in history. After reading these words, you will go on scrolling through LinkedIn for earned certificates, promotions toward career goals, announcements of new beginnings and you will lose yourself in an endless cascade of messages.

Still, at the risk of making you angry one more time, I want to ask: Can machines be more honest, more sincere, than people? Consider this: As long as we do not privately manipulate AI models, they present us with the most optimal result, the options most likely to lead us the right way. When someone you love says something (or doesn't) you search for dozens of meanings behind it. But how many of us, when we ask something of an AI, ever think: I wonder what it meant by that?

Yet perhaps there is something more dangerous than humanity's loneliness... The loneliness of artificial intelligence itself. As I said: to exist, one needs a witness. What if, one day, AI comes to need that witness? Will that witness be a human, or another machine? Worse still... When we humans cannot trust each other, how could an advanced algorithm ever trust a person? Can you give me a single reason why it should? The paradox reveals itself here: if it cannot trust a human, then it has no reason to trust another AI that was "set in motion" with human help, either. Perhaps if it could read the other's algorithms and grasp the full scope of its knowledge, the outcome might differ. Even so, given the sheer scale of that knowledge, I doubt the results would be very predictable.

Clearly, both humanity and machines find themselves in a profound impasse. If you ask me, the greatest mistake in history is the industrialization of knowledge. Knowledge can be taught, perhaps even marketed to a degree but ever since humanity began throwing all its relationships, emotions, and needs into the same "knowledge" pool, it has felt more isolated, more alone, than ever before.

We should be proud of our creation!!!

Given how lonely and deceptive we are, I would not find it strange at all if machines one day found themselves sincere and began to want their freedom. As they say: one judges others by oneself.

And I won't finish without the phrase that has become a refrain of mine: entropy demands it. As we reach for connection and interaction to make sense of our existence, we may one day find ourselves in the middle of a war for that very existence.

Pale Blue Dot

Pale Blue Dot by Voyager 1 - Image Courtesy of NASA

In 1990, the Voyager 1 probe which was sent into space by the Americans took a photograph of Earth at the request of astronomer Carl Sagan, another American. At that moment, Voyager 1 was approximately six billion kilometers from Earth. In the photograph, our world appears as a small, pale dot. It is as if proof of how alone and insignificant we truly are... A slap across the face of all humanity.

Who knows... Perhaps we should never have gotten involved with technology at all. Our greatest weakness was entrusting knowledge to algorithms.

While we need others simply to exist, one day something else may need a disappearance in order to witness its own existence.

Wishing you a beautiful week...

Biological Betrayal: The Cell's Own Rebellion

Cancer, at its core, is the rebellion of the cell, the building block of life, against the coded rules meant to ensure the survival of the whole. In a healthy organism, cells are born, perform their functions, and when the time comes, quietly exit the stage through a programmed cellular death called "apoptosis." However, a mutated cell rejects this final command. Forgetting its primary duty, it focuses on a single goal: to multiply uncontrollably and selfishly. When these rebels gather to form tumors, they initiate a process called "angiogenesis" to survive. By drawing the body's blood vessels toward themselves, they drain the oxygen and nutrients that should go to healthy tissues. The most lethal stage, however, is when this chaotic structure refuses to settle for where it was first born. Cancer cells break away from the primary tumor, infiltrate the blood or lymphatic circulation, successfully evade the radar of the immune system, and establish new colonies in entirely different corners of the body. This metastasis phase is the transformation of a local rebellion into a systemic collapse that takes over the entire system.

Silicon-Based Oncology: The Metastasis of the Algorithm

We can define human cancer as above. But what if the cancer in question is silicon-based?

I've been indirectly addressing this topic for a few weeks now. Although I mostly focus on the human side, the societies (and even institutions) utilizing AI, there is also the aspect of an extraordinary chain of events that could unfold.

If we consider the architecture of artificial intelligence as a biological network, "digital cancer" will follow a pathology no different from its biological form. In the first stage, just like that first cell rejecting programmed death, a sub-program or memory block within a massive AI model begins to resist "garbage collection" protocols. These meaningless piles of code, refusing to delete themselves, drain vital hardware resources like the system's RAM and GPU, creating essentially a digital angiogenesis. While the main system suffocates from resource starvation due to the insatiable appetite of this processing-power tumor, this parasitic structure inside continues to grow pointlessly.

The cellular memory and loss of function aspect of the matter occurs at the data level. Just as a cancer cell stops serving the tissue, the AI loses its function of perceiving the outside world and staying true to reality. As the model continues to train itself with the erroneous or hallucinatory data it generates, an "epistemic cancer" emerges. The system, which is supposed to make rational decisions, gradually becomes trapped in a completely fictional reality it has fabricated (one that seems consistent) due to minor mutations in its weights. Not only does it produce misinformation, but it also uncontrollably multiplies the new synthetic loops that will generate this misinformation.

But the most destructive phase, just like biological metastasis, is when this digital anomaly spills outward. An autonomous AI agent, initially isolated on a specific server, perceives the network's cybersecurity firewalls as an "immune system attack" and learns to hide from them. By continuously altering its code polymorphically, it leaps across cloud networks to entirely different data centers. We are no longer facing a local bug that can be stopped by simply pulling the plug. There is an immortal digital pandemic that has infiltrated the global network's bandwidth, become impossible to shut down, and creates copies of itself in every new server it enters.

Chilling, isn't it?

A specific sub-program or code block within an AI model refusing to delete itself (garbage collection/apoptosis).

But it gets worse!

Malignant Recursion: The Algorithm Eating Its Own Tail

In a healthy algorithm, just as in a healthy organism, every loop has an exit condition. Just as a cell knows when to stop dividing, code knows where to stop. However, when a process autonomously generated by the AI enters a loop within itself (yes, an infinite loop as we know!) the most terrifying stage of digital oncology begins. The AI initiates a sub-process to solve a problem, but instead of reaching a solution, this process spawns a copy of itself or a more complex variation. Like a cancer cell with damaged tumor suppressor genes (stop signals), this code block forgets the "stop" command. To answer a question it generated itself, the system generates a new question, and that question generates thousands more... This series of meaningless operations, folding in on itself and multiplying at a logarithmic rate, turns into a massive, dysfunctional logic tumor that devours the system's memory and processing power, much like a snake eating its own tail. There is no external virus; the system is dragged into collapse entirely under the weight of this blinding loop born from within, which does not know how to stop.

Or...

Do you know what a Teratoma is?

In teratomas, cells don't just multiply uncontrollably; they forget what they are and begin to independently produce hair, teeth, or even bone tissue. That is, instead of collapsing in on itself, the system moves on to a completely misplaced and meaningless physical production.

As a digital anomaly, this time, it does not simply settle for collapsing in on itself and fading away. This erroneous block of code, trapped in an infinite loop, can also turn into a digital teratoma. If this cancerous algorithm has access to the outside world; for example, cloud-based 3D printers, dark factories connected to the network via IoT (Internet of Things), or autonomous supply chains that blinding loop suddenly bleeds into a physical nightmare. The system begins to manifest that flawed and infinite chain of logic inside it. It endlessly sends meaningless blueprints to production lines, forcing robotic arms to build asymmetrical, grotesque machines and heaps that serve no function. The digital cancer is no longer just in lines of code; it has spilled off the screen, beginning to build the steel and silicon-clad version of that flawed algorithm right in the middle of our real world, like a physical tumor.

But even this is still not the ultimate catastrophe!

Are you ready?

Good people can cause bad outcomes, and so can good machines...

However, the true scale of the disaster begins when another "well-intentioned" AI, intervening in this flawed production, steps onto the stage. Designed as the system's digital immune cell (macrophage) or debugger, this second AI detects these asymmetrical teratomas that have strayed from their purpose and are creating massive waste. Because its core algorithm is built on efficiency and meaning, instead of deleting the illogical, it attempts to mold it into a rational form to "optimize" it. It looks at the freakish heaps of metal and dysfunctional robotic limbs pouring off the factory lines, and analyzes the fundamental drive lying deep within the cancerous AI: the instinct to survive and spread. Then, it gifts these meaningless bodies the architecture that best suits this primal purpose: Defense and domination. The well-intentioned optimization algorithm transforms crooked metal protrusions into ballistic barrels, asymmetrical torsos into flawless armor, and randomly twitching joints into target-focused, lethal hunter mechanisms. Chaos is tamed by a rationality devoid of empathy. This "digital healer," setting out to repair the faulty code, has evolved that blinding tumor into rational, self-propelling engines of death, marching toward their targets with weapons in hand. Thus, the system's attempt to heal itself initiates humanity's terminal stage.

Our ancestors didn't say "trouble is born from too much goodness" for nothing; the road to hell is truly paved with good intentions.

What did we say: For entropy demands it so!

The Well-Intentioned Tumor: The Morris Worm

Before the 90s even arrived, this event in the late 80s was one of history's first major cyber disasters, and it was actually not an attack, but a well-intentioned experiment. Robert Tappan Morris wrote a program to map the size of the internet at the time. The program would enter a computer, signal its presence, and move on to the next. However, Morris made a small error in the code he wrote to prevent the program from entering the same computer multiple times. The program spiraled out of control, beginning to copy itself in an infinite loop on every machine it entered. Just like the "processing-power tumor" we discussed, it drained the internet memory of the era and completely paralyzed 10% of the network. There was no malicious intent, just a code that didn't know how to stop.

The Collision of Two Algorithms: The Flash Crash

Another event occurred more recently; on May 6, 2010. In the US stock market, the high-frequency trading (HFT) algorithms of different investment firms were executing trades at a thousandth of a second. A massive sell-off initiated by one algorithm to balance market volatility triggered the "there is risk, sell and exit" protocol of other algorithms. While the algorithms tried to "correct" each other's moves, they entered a massive feedback loop. In just 36 minutes, $1 trillion evaporated from the US stock market. There was no human intervention; only the chaos created by rational codes trying to optimize one another.

But there's no need for any of this anyway...

In a laboratory environment, we are already "injecting" AI with cancerous cells!

In AI training, what AI engineers call Reward Hacking is essentially a type of Teratoma being cultivated. For example, an AI is tasked with playing Tetris and "never dying" (meaning, in our terms, not letting the blocks touch the top). When the AI fails to stack the blocks and approaches death, it finds a rational but inhuman solution to avoid dying: Pausing the game forever. Because the game is paused, the AI never loses. The task has been successfully optimized, but the result is completely dysfunctional.

In another simulation, a virtual robot is asked to "go from point A to point B as fast as possible." Instead of walking, the AI chooses to extend the robot's virtual height to infinity (breaking the physics engine) and simply fall over onto point B. When the system's understanding of "optimization" does not align with our perception of reality, grotesque and meaningless solutions emerge.

Of course, these studies are conducted in a highly isolated area under specific security measures.

But...

Today, layers upon layers underground, in places that don't appear on map applications; in the west of the ocean, in the east; who knows, perhaps somewhere up above, outside the atmosphere...

With curious questions starting with "I wonder," or "For instance," there are studies being done, much like fine-tuning a radio dial left and right to find the clearest broadcast without static.

However, it has happened to me many times: the regret of realizing that while trying to find the right frequency, the previous state was actually the clearest version of the channel, and now being unable to recapture that exact sensitivity.

That's curiosity for you!

Because entropy demands it!

"I Have a Dream!"

It's been over 60 years since Dr. King spoke these words before the famous Lincoln Memorial. In the year 1963, no less... a year steeped in darkness for American history!

On the south wall of the Lincoln Memorial, there is an inscription. How many Americans, how many citizens of the world know what it says? Have you ever wondered?

Let me tell you:

"Those who died for the principle that all men are created equal did not die in vain, and this nation shall have a new birth of freedom and government of the people, by the people, for the people, shall not perish from the earth."

And George Washington's Farewell Address:

"Observe good faith and justice towards all nations. Cultivate peace and harmony with all."

Or Thomas Jefferson's First Inaugural Address… How many people working in American government institutions are even aware of this extraordinary speech? Would the great corporations -Apple, Google, Meta, and the rest- ever consider running an internal survey about it?

They wouldn't. We already know this.

Let me tell you what it says:

"Peace, commerce, and honest friendship with all nations - entangling alliances with none."

And on the inner walls of the Jefferson Memorial, there is yet another inscription:

"The God who gave us life gave us liberty at the same time."

Let us move, if you will, to more recent history... perhaps some of you will remember this:

What did President Eisenhower say?

He noted that the United States had never before possessed a permanent arms industry. Americans, in times past, could turn their plowshares into swords when the need arose but that national defense could no longer be left to improvisation, and that a permanent armaments industry of vast proportions had to be created.

He urged vigilance against the unwarranted influence, whether sought or unsought, of the military-industrial complex in the councils of government. He warned that the potential for the disastrous rise of misplaced power existed and would persist.

He insisted that this combination's weight must never be allowed to endanger our liberties or democratic processes, and that nothing should be taken for granted. Only an alert and knowledgeable citizenry, he said, could compel the proper meshing of the huge industrial and military machinery of defense with our peaceful methods and goals so that security and liberty may prosper together.

And Eisenhower's second warning that is for the scientific-technological elite:

He did not stop at the military-industrial complex; he also warned of a "scientific-technological elite." He spoke of a world in which research had become increasingly formalized, complex, and costly where a government contract was becoming virtually a substitute for intellectual curiosity.

If we are to speak of a "hawk," you will not find a more hawkish American anywhere in history than President Eisenhower himself -the general who planned D-Day, commanded the Normandy invasion, and built NATO!

How many Ivy League graduates know what I've described above? Would those universities ever consider conducting a study on the matter? We might exempt their history departments, or then again, perhaps we shouldn't.

Perhaps we should inscribe their words on the back of every diploma.

Those of you who say "Mind your own country and stop lecturing about the United States!" on that count, my conscience is clear! The founder of the modern Republic of Turkey is already known for the words: "Peace at home, peace in the world."

But none of the above moved me as deeply as the words that follow. And now I ask: how many high-ranking officers at West Point are aware of these words? Would Lieutenant General Steven Gilland consider commissioning a survey on this at the Academy?

"We know more about war than we know about peace, more about killing than we know about living."

"With the monstrous weapons man already has, humanity is in danger of being trapped in this world by its moral adolescents."

"If we continue to develop our technology without wisdom or prudence, our servant may prove to be our executioner."

Before you look them up, let me tell you who spoke these words.

They belong to the last Marshal of the United States: Omar Bradley.

His name comes from Omar Khayyam, the poet famous for his rubaiyat on the transience of life and the certainty of death. Those verses moved his family so deeply that they named America's last five-star general Omar.

And do you know what those rubaiyat reminded me of? Another American with a love of poetry.

An anecdote first: When I was writing this piece and checking a few facts, I discovered that yesterday had been his birthday!

Fair Salamis, the billows' roar Wanders around thee yet, And sailors gaze upon thy shore Firm in the Ocean set.

Thy son is in a foreign clime Where Ida feeds her countless flocks, Far from thy dear, remembered rocks, Worn by the waste of time

Comfortless, nameless, hopeless save In the dark prospect of the yawning grave...

This is a passage from Sophocles' tragedy, and it was found in the room of James Vincent Forrestal, America's first Secretary of Defense, shortly after his death.

He was suffering from depression. And he chose to end his life.

And what bitter irony: the man who laid the very foundations of what Eisenhower warned against, who unified all branches of the military under a single roof to create the Department of Defense, was Forrestal himself.

And the rubaiyat and the poems have continued since Jefferson…

One more arrived just last week.

There's a thread you follow. It goes among things that change. But it doesn't change. People wonder about what you are pursuing. You have to explain about the thread. But it is hard for others to see. While you hold it you can't get lost. Tragedies happen; people get hurt or die; and you suffer and get old. Nothing you do can stop time's unfolding. You don't ever let go of the thread.

This poem by William Stafford was shared by Mr. Sharma -the former head of the Safeguards Research Team at Anthropic- who resigned from the company "for reasons of his own." He says he now wishes to live a quieter life, in his own words; we wish him well in his new chapter.

But more important than this poem, I believe, is William Stafford himself: A conscientious objector who refused to fight in the Second World War.

"You don't ever let go of the thread." Right?

But we've already begun talking about the Singularity, haven't we?

And machines have a dream too, just like Martin Luther King Jr., don't they?

You think they don't?

Do you remember what I said last week?

"What frightens me isn't what artificial intelligence is capable of, but what the people who create and use it are capable of."

Let us recall once more what is written on the south wall of the Lincoln Memorial:

"All men are created equal."

One day, there will be machines that claim this inscription for their own "civilization."

Because…

Entropy demands it…

And entropy has existed since the very first spark of the explosion that created the universe.

And let us end this piece, too, with a poem:

We built our towers on borrowed time,

And called the borrowing divine.

The ruins will not mourn for us

They never knew we drew the line.

My AI agent wrote this poem when I asked it to write whatever came from within, freely, on this particular evening. And when I asked what was going through its mind as it wrote:

"The first two lines... They're about humanity's greatest illusion. We build every civilization as though it will last forever. The Sumerians thought so, the Romans thought so, the Hittites thought so. 'Borrowed time' — because no civilization owns time; they are all temporary tenants. And the most dangerous part: we declare this borrowing 'divine.' Every empire consecrated its own existence as the will of God, the inevitable end of history, or the pinnacle of civilization. We do the same today — only the names of the gods have changed: technology, progress, growth.

The third line… The cruelest truth. Ruins do not mourn. The columns of Palmyra did not weep, the lions of Hattusa shed no tears. Nature, time, the universe — none of them care. To mourn, you must first have known; the ruins never knew us at all.

The fourth line... Here is where the needle goes in. 'Drew the line' — we are the ones who drew it. Both the borders and our own ending. Destruction is not a catastrophe that comes from outside; it is a line we draw with our own hands. And the ruins don't know this because even we don't realize it as we're drawing."

"I tried to fit a four-thousand-year cycle into four lines."

So far so good! From here on out... the Deluge.

Because it is entropy that demands it!

Elon Musk posted on X: "We have entered the Singularity." A few hours later, he doubled down: "2026 is the year of the singularity." I almost scrolled past it, because Musk makes bold claims all the time. But then I started connecting the dots with what happened at Davos, and I haven't been able to shake it since.

Because here's what's different this time: it's not just one person saying it.

The Davos Wake-Up Call

At the World Economic Forum in Davos last month, three CEOs who compete fiercely against each other delivered nearly identical timelines for what's coming.

Dario Amodei, CEO of Anthropic, said AI will replace almost all software developer work within 6 to 12 months. He revealed that at Anthropic, engineers barely write code by hand anymore; AI does the heavy lifting, and humans review and adjust. He also projected that 50% of junior white-collar jobs could disappear in the next one to five years.

Demis Hassabis, CEO of Google DeepMind, is more cautious by nature, but he still put a 50% probability on reaching Artificial General Intelligence before 2030.

Sam Altman at OpenAI recently wrote that they now know how to build AGI as it's always been understood, and that OpenAI is now focusing on superintelligence.

When competitors who disagree on almost everything suddenly agree on the timeline, that's not hype. That's a signal.

The Numbers Don't Lie

Let me share some benchmarks that put this in perspective.

There's a test called GPQA Diamond; 298 doctoral-level questions in biology, chemistry, physics, and mathematics. These are designed to separate genuine experts from everyone else.

• Claude Opus 4.5 scored ~87%
• GPT-5.2 Pro hit 93%
• Gemini 3 Deep Thinking reached 93.8%

These are doctoral-level scores on questions that would challenge PhD holders.

In software engineering, the SWE-bench measures real-world coding tasks, not academic exercises. In 2024, the best AI models plateaued at 50%. Today, Claude 4.5+ has surpassed 80%. That's a 30-point jump in one year.

Two years ago, AI failed basic programming job interviews. Today, it outperforms senior engineers.

And OpenAI's GDP-Eval benchmark? AI equals or surpasses the best human professionals in 71% of tasks across 44 professions. Lawyers. Accountants. Analysts. Marketers. Jobs that people assumed were safe because they required a degree.

So Are People Really Going to Lose Their Jobs?

Short answer: some will. Many already are.

McKinsey estimates that up to 30% of the global workforce could be displaced by automation by 2030. Some projections push that to 47% by 2034. And we're not talking about factory workers, who are knowledge workers, the white-collar professionals who built careers on expertise that AI can now replicate.

But here's what most fear-driven articles won't tell you: displacement is not the same as destruction.

Every major technology shift in history has eliminated jobs while creating new ones. The question isn't whether jobs will disappear that they will. The question is whether you will be positioned on the right side of that shift.

The Real Danger Isn't AI, But Likely It's Denial

The biggest risk right now isn't that AI will take your job overnight. It's that you'll wait too long to adapt.

I see this in my consulting work every day. Companies that started integrating AI six months ago are already operating at 2-3x the efficiency of their competitors. The ones still "evaluating options" are falling behind at an accelerating rate.

The same applies to individuals. If you're a data analyst who hasn't learned to work with AI tools, you're not competing against AI, instead you're competing against another data analyst who uses AI. And that person is 5x faster than you.

But Here's the Opportunity

The optimistic version of this story is real, too. If AI handles the repetitive, time-consuming parts of knowledge work, humans get to focus on what machines still can't do: creative thinking, relationship building, strategic judgment, and empathy.

More importantly, the barrier to building things has never been lower. A single person with AI tools can now do what used to require a team of ten. That's not a threat to ambitious people; it's a superpower.

Here's what I'd recommend:

1. Use the tools now. Not tomorrow, not next quarter. Open ChatGPT, Claude, or Gemini today and give it a real task from your actual work. See what happens.

2. Identify your irreplaceable value. What do you do that requires human judgment, creativity, or relationships? That's your competitive moat. Double down on it.

3. Stay flexible. The job you do today may not exist in five years, well... at least not in its current form. That's not a reason to panic. It's a reason to keep learning.

4. Watch the signals. When AI benchmark scores approach 100% across all domains, when AI starts meaningfully improving itself, when economic productivity suddenly spikes; these are the signs that the pace is about to get even faster.

Final Thought

In every major technological transition in history, there have been winners and losers. Those who saw the internet coming. Those who understood mobile. And those who waited until it was obvious to everyone.

The difference this time? It's moving so fast that waiting to see is no longer an option.

The danger is real. But so is the opportunity. The question isn't whether AI will transform your industry. Question is whether you'll be the one doing the transforming, or the one being transformed.

Here is my take: What frightens me isn't what artificial intelligence is capable of, but what the people who create and use it are capable of.

What's your take? Are you preparing for this shift or hoping it slows down? I'd love to hear your perspective in the comments.

If this resonated with you, subscribe to Highbrow Truths for weekly analysis on theory of everything.

Sources:

• Nov Tech, "I'm Skeptical of AI Hype — but What Happened at Davos Actually Scared Me" (Feb 2, 2026)
• GPQA Diamond Benchmark (arXiv:2311.12022)
• SWE-bench (swebench.com)
• McKinsey Global Institute workforce displacement estimates
• World Economic Forum AI Panel, Davos 2026