MARC ANDREESSEN AND THE GODS OF POISONED SILICON

The Most Dangerous Thing About AI Is That It’s Being Built by the Least Human People of All

Marc Andreessen broke the internet into a thousand pieces two days ago (60,000+ posts mentioning him, last time I checked), by saying something on David Senra’s podcast that was as astonishing as it was idiotic. Or as disruptive as it was alarming. Senra asked Andreessen, one of the managing partners of the famed venture capital firm Andreessen Horowitz, whether he practiced introspection. 

“Zero. As little as possible. All you need to do is keep moving forward. Go!”

He went on to claim that the great men of history had no introspection either (introspection being the process of self-observation and internal analysis through which a person examines their own thoughts, emotions, memories, and behaviors), and that the whole thing is a problem both at work and at home because it leaves people spinning their wheels instead of moving forward. Then, remarkably, Senra (the interviewer) chimed in to say that after reading 400+ biographies of entrepreneurs and influential figures, he never found any evidence of introspection in them. That Sam Walton didn’t wake up thinking about his inner self; he just built more Walmarts. To which Andreessen replied that this whole culture of self-examination traces back to Vienna, to Freud, to the 1920s, a kind of guilt imported from Europe that infected people with the urge to stare at their own navels instead of conquering the world. 

He also said that “if you go back 400 years, it wouldn’t have occurred to anyone to be introspective.” 

But then he said something even bolder:

“There is no inner self. You’re looking for an imaginary concept. Period.”

And then he hammered a few crude nails into the coffin, declaring that introspection is neurosis.

Look, setting Senra aside for a moment (who may have been suffering from an acute case of sycophancy), we’re not talking about some crank here. Marc Andreessen is a highly influential billionaire who has achieved enormous success, who shapes or outright defines the future of entire industries, and who has helped Silicon Valley evolve and thrive. His opinions carry something close to the weight of law. So what he believes about human nature doesn’t stay confined to a podcast interview; it becomes the governing philosophy that steers thousands of entrepreneurs and trillions of dollars toward the technologies that will organize how we live, what we eat, what we think, how we love, and how we die. He’s not a fool. So the most reasonable thing to do is think carefully about what he actually said.

That’s exactly what I’ve been doing for the past two days, ever since I first heard him say all this, despite my nervous laughter while watching the interview. In any case, drawing on nearly three decades of experience analyzing human behavior and psychology, performance, and entrepreneurship, and also drawing on the hundreds of thoughtful responses I read from respected people who pushed back and challenged his views, I’ve come to the conclusion that this is a complex and delicate issue. A dangerous one. But also a timely one for you, someone who isn’t a billionaire but still wants to make a dent in the world. Which means you need to understand what’s at stake. Keep reading.

WHERE MR. MARC IS RIGHT

Let’s be fair: 

He’s right about exactly one thing. Just one. The problem is that he uses that single truth as a springboard to launch himself into a putrid abyss, and on the way down, he tries to drag all of humanity with him.

Because there is a form of looking inward that genuinely is poison. It’s called rumination. It’s the mental loop that goes nowhere: replaying the past without learning from it, punishing yourself without changing, picking at wounds without healing them. The science on this is unambiguous. Rumination is associated with depression, cognitive distortions, worse performance, and a sustained erosion of wellbeing. On that specific point, Andreessen is right. That kind of sterile introspection does destroy.

But here’s where things go sideways. Because Andreessen takes that partial truth and uses it like a scalpel to amputate something essential from our humanity.

THE LEAP

There’s a world of difference between saying:

“Paralyzing rumination is destructive and you need to act even without certainty”

…and saying:

“There is no inner self. You’re a fifteen-second context window with a goldfish’s memory. Introspection is neurosis dressed up in philosophical clothing.”

The first statement is psychologically precise. The second is an abusive philosophical reduction disguised as pragmatism. It’s the equivalent of turning medicine into poison.

The American philosopher Dreyfus explained this well when he argued that human intelligence is not the formal manipulation of symbols. It is embodied. It is situated. It carries context, a body, history, and vulnerability. Not because memory is perfect, but because from its imperfection you cannot conclude that human beings are reducible to a computational metaphor.

The fact that your memory is fallible does not mean you have no self. The fact that your introspection can be imprecise does not mean it should be abolished. The fact that moving forward is necessary does not mean that self-examination is cowardice. Or stupidity.

Andreessen makes that leap as though it were obvious and necessary, as though the distance between “act with courage” and “the inner self is a fiction” were just a single natural step. But it isn’t. It’s a cliff. And at the bottom, something very serious is waiting with its jaws open.

THE PEOPLE WHO ACTUALLY USED INTROSPECTION TO CHANGE THE WORLD DID IT BETTER

Andreessen says the great men of history didn’t examine themselves. They just built.

The problem is that history flatly contradicts him. This is almost certainly the most embarrassingly wrong thing this human money-printing machine has ever said.

• Because apparently he never heard that Marcus Aurelius ruled the largest and most powerful Roman Empire of his era, and every night before going to sleep, he wrote in his journal. Not about his victories. Not about his conquest strategies. He wrote about his own failures, his impulses, his tendency toward anger, his fear of what others thought of him. Those writings became the Meditations, one of the most influential texts in Western philosophy. An emperor who built an entire world from the inside out.

• Abraham Lincoln navigated the most devastating civil war in his nation’s history while living with a deep, well-documented melancholy. He never hid it, and in fact he examined it unflinchingly. And that capacity to hold emotional complexity without running from it was, according to many historians, precisely what allowed him to make decisions of extraordinary political maturity in the nation’s darkest moments.

• Nelson Mandela spent twenty-seven years in prison. Twenty-seven years that would have turned any mind without introspection into pure hatred. Instead, he emerged with a moral clarity that transformed an entire nation. He himself described those years as a period of deep self-examination. Not paralysis. Forging.

• Steve Jobs, Silicon Valley’s own patron saint, by the way, practiced Zen meditation rigorously for decades. He studied deeply the meaning of his own life, immersed himself in contemplative traditions, and explained many times that his capacity for introspection allowed him to see things others couldn’t. The design of Apple, his obsession with user experience, his ability to bridge technology and human emotion... all of it was born from a deep practice of presence and self-knowledge.

• Leonardo da Vinci himself used introspection to fuel his creativity and performance. He told his students: “Leave your family and friends and go into the mountains and valleys, go out into the fields. While you are alone, you are entirely your own master.” His more than 6,000 pages of notebooks are the physical proof of a life turned inward. In them, Leonardo didn’t just sketch machines and bodies; he gave himself instructions, questioned himself, and corrected himself.

  To Mr. Andreessen’s likely surprise, biographers like Giorgio Vasari recount that Leonardo could spend entire days standing before The Last Supper without touching a brush, simply looking and reflecting. When challenged on this, he explained that geniuses “work most when they work least,” because they are cooking ideas from within.

The list goes on and on: Gandhi. Einstein. Simone Weil. Churchill. Ray Dalio. George Soros, who made transcendent investment decisions based on signals from his own back pain...

But here’s what’s truly important: the difference between those who merely built empires and those who left something worth preserving is not the sheer volume of their execution. It’s the depth from which they acted. It’s their... yes, that. Their introspective depth.

But Andreessen, after spending decades surrounded by entrepreneurs, has concluded that the best ones have zero introspection. Perhaps because his goldfish-bowl view of human cognition, which he seems to believe is the only model available to all of us, prevented him from seeing anything beyond what he expected to see, thanks to his rather enormous confirmation bias.

THE THEORY TAKES SHAPE

This is where we need to stop talking about Andreessen as an individual and start talking about what he represents.

Because the problem isn’t one man. It’s a pattern I see in many of the new power players coming out of Silicon Valley, people whose robotic pragmatism has reshaped them so thoroughly that I’m almost ready to believe even their bathroom breaks run on a silicon algorithm.

The pattern goes like this: when a person spends enough time immersed in systems of optimization, computation, scaling, AI, velocity, billions of dollars, and KPIs governing every move, something begins to shift in the way they interpret reality. Not all at once. Almost certainly not consciously. But in the way sand dissolves into a wave: slowly, irreversibly, until you can no longer tell what it was from what it is.

They begin to see everything as input and output. Emotion becomes noise, an obstacle. Pausing becomes inefficiency. Doubt becomes friction. Other people become functional nodes inside a system. And the inner self becomes a romantic fable that slows down the process.

This has a philosophical name. Horkheimer, the German sociologist, psychologist, and philosopher, called it instrumental reason: a form of rationality that is supremely skilled at optimizing means but has lost the ability to ask

what for. An intelligence that is prodigiously effective at the “how” and profoundly blind to what actually matters.

Heidegger went further and called it the danger of enframing: the tendency of technological logic to convert everything into a standing reserve, a resource available for use. Not just nature. People too. Time too. The self too. The future too.

And Hartmut Rosa, the philosopher and sociologist, adds something that stings especially today: a civilization organized around permanent acceleration, compulsive innovation, and “go, keep moving, go” loses something he calls resonance. The living, receptive, transformative relationship with the world. With oneself. With others. With time. 

This is how you can be enormously efficient and profoundly out of balance at the same time. In fact, that’s exactly what we’re watching unfold.

And I can’t help it, but when I watch Sam Altman in his interviews, he comes across to me like a mannequin that’s been fitted with skin and hair. He speaks from inside a rational framework so solid, so tightly sealed, that it blocks the expressions and gestures that tend to come naturally to human beings. I bring him up because in one interview I heard him say, with what seemed like genuine satisfaction, that fortunately he doesn’t have much emotional impulsivity.

And like him, I’ve noticed for a while now that many others in that circle of the now-poisoned valley layer the unnatural, the robotic, the measurable, and the computational over everything that smells of flesh and blood. And if you read between the lines of what they write and say, you find a subtle but unmistakable contempt for the qualitative, the unmeasurable, the emotional. For the very things that make us most human.

THE THREAT TAKING SHAPE 

To be clear, these people aren’t mutating or conspiring. These men don’t wake up wanting to harm people. But the danger can be as subtle as it is real, because it’s a gradual drift of the mental frame toward an instrumental logic that impoverishes the vision of what it means to be human, and the person it’s happening to doesn’t even notice.

The argument goes like this: prolonged immersion in high-abstraction, high-optimization, high-velocity technological environments induces in certain individuals a cognitive and moral drift, through which they begin to interpret human reality using the same categories they use to design machines or print money. And in doing so, they reduce their sensitivity to irreducible dimensions of life: interiority, vulnerability, moral ambiguity, empathy, presence, intrinsic dignity.

They don’t become machines (I think). But they grow accustomed to a framework in which only what can be measured, modeled, predicted, and scaled seems real. And within that framework, the inner self is noise. Introspection is friction. Pausing is waste. Emotion becomes a system error.

And that’s where the real danger lives.

Not for their software. For civilization.

WHO’S ACTUALLY HOLDING THE WHEEL?

We need to be precise here, because this is not a romantic complaint against technology.

By definition, technology does not dehumanize. The wheel didn’t dehumanize anyone. Neither did the printing press. Electricity did so even less. The problem is never the tool. The problem is what vision of humanity guides the person building it.

Because every powerful technology carries an implicit idea of what it means to be human. The AI that decides what content you see carries an idea of what you deserve to see. The algorithm that determines your credit score carries an idea of what you’re worth. The platform that organizes your relationships carries an idea of what it means to connect with another person.

And if that idea of what it means to be human is built by people who have concluded that the inner self is a fiction, that memory is mostly unreliable, that we are fifteen-second context windows, that self-knowledge is stupidity (and neurosis)... what kind of world are they designing for us?

Social psychology already has a name for what happens when people are perceived as objects, automatons, or resources: mechanistic dehumanization. Nick Haslam documented how this form of perception erodes empathy, moral responsibility, and the capacity to feel the harm being inflicted on others.

And there’s something even more unsettling: organizational research shows that in highly algorithmized, objectifying work environments, people begin to perceive themselves in more instrumental terms. Objectification doesn’t just change how you see others. It changes how you see yourself.

MIT professor Sherry Turkle has spent years documenting what happens when we grow accustomed to expecting more from technology and less from other people. Deep human conversation, she argues, is the muscle of empathy. And environments that reward speed, abstraction, and transaction slowly erode it. Slowly. Quietly. Without anyone noticing.

Until we can no longer remember what it felt like to have it.

THE MISTAKE HE’S ALREADY MADE

Andreessen commits a classic philosophical error that in academia is called the fallacy of composition: because certain empire builders were successful with little introspection, he concludes that introspection is irrelevant or harmful to human beings in general.

But that reasoning has a glaring flaw.

If you obsessively study empire builders, you will naturally find profiles marked by high action, low hesitation, aggressive simplification, and relentless forward motion. That selection bias does not prove that those men represent the most complete form of humanity. It only proves that that particular profile works well for certain kinds of conquest within certain kinds of systems. In other words, it can help you win. Sure. But that kind of winning reduces our humanity to the most brutish side of our animal nature.

Take Mark Zuckerberg, one of the anointed gods of Silicon Olympus. He built the largest communication platform in the history of humanity. Two billion people. A technical and entrepreneurial feat without precedent, allowing us to connect and socialize in entirely new ways. Good? Absolutely. But the Cambridge Analytica scandal revealed that Meta’s engineers knew the data of eighty-seven million people had been extracted and used to manipulate elections, and nobody stopped to ask whether that was okay. And then something even more serious came to light: Meta’s own internal documents showed the company knew Instagram was causing documented harm to the mental health of teenagers, especially girls, and chose not to change the algorithm because user engagement was too valuable. They knew. They had the data. They kept going. Now spend a moment thinking about how the thing Meta created is shaping the adults of the future; not thousands of them. Billions.

But now look at the other side of the coin.

Warren Buffett is ninety-four years old and still one of the wealthiest people on the planet. But what sets Buffett apart isn’t just the size of his fortune; it’s the mental architecture from which he built it. Buffett spends hours every day reading, thinking, and examining his own mistakes with brutal honesty. He has publicly documented his failures, his biases, his moments of miscalibrated greed. His partner Charlie Munger described him for decades as a man obsessed not just with understanding the market, but with understanding his own limitations as an observer of the market. That distinction is not trivial: knowing that your judgments are contaminated by who you are requires exactly the kind of introspection Andreessen dismisses. And in 2006, Buffett did something no startup optimization algorithm would ever have recommended: he donated 99% of his fortune. Why? Because, in his own words, he had examined his life with enough depth to conclude that accumulating wealth beyond a certain point added nothing worth having. Mr. Andreessen, do you know what that is? Pure self-knowledge. Functional introspection.

And George Soros? The one who makes transcendent decisions based on what his back is telling him? Well, having donated more than $32 billion to the Open Society Foundations (64% of his net worth), he has been called the “most generous donor” in the world relative to net worth. And the back-pain thing, which might seem like an odd animal quirk, takes on serious significance once you understand that Soros constantly monitors how his own biases are contaminating his market analysis. When there’s a contradiction between his theory (what he thinks is true) and reality (what’s actually happening), his body reacts. That interoceptive signal (the back pain) alerts him that his introspection has detected a flaw in the system. As he himself put it:

“I am critical of myself to the point of cruelty... that is the only way to survive in a market that does not forgive arrogance.”

So: what model of greatness do we want to build into artificial intelligence, synthetic biology, and the algorithmic governance of societies?

Because if it’s the first model, we have a very serious problem. And one bigger than anything we’ve faced before, because AI grants incalculable power that, in the wrong hands, can produce a social, geopolitical, military, economic, cultural, and even existential catastrophe.

WHAT THE SCIENCE SAYS, CLEARLY 

Contemporary neuroscience also has standing to weigh in on this. And unsurprisingly, it becomes yet another field that categorically dismantles Andreessen’s argument: useful introspection exists, it works, and it produces measurable outcomes.

Lisa Feldman Barrett and other rigorous neuroscientists and psychologists have documented that developing what is called emotional granularity, the ability to distinguish internal states with precision, improves emotional regulation, improves decision-making, and reduces maladaptive behaviors. This isn’t philosophy. These are studies. This is data. This is introspection.

The critical distinction science draws is between rumination and adaptive reflection, as I noted earlier. Rumination is indeed circular, sterile, self-focused, and negative. But adaptive reflection involves psychological distance, genuine curiosity, and a learning orientation. One destroys. The other builds.

The problem isn’t looking inward. The problem is how you look and what you’re looking for. And that, precisely that, requires a level of self-knowledge that Andreessen dismisses entirely.

Understanding without acting stagnates. Acting without understanding perpetuates patterns with no deep value. The most solid approach, according to the evidence, is using introspection in service of action. Not as a substitute for it. As its compass.

THE PRACTICAL STAKES

All of this would have limited impact if Andreessen were just an eccentric thinker with provocative opinions. But he’s not. 

That’s why my reflections here amount to at least one urgent call to action: the teams building AI, platforms, and algorithmic systems desperately need something Silicon Valley culture systematically despises: training in the humanities, ethics, philosophy, and the psychology of mind and human development. More than an academic ornament, it needs to be an operational requirement. Because designing systems that affect billions of people without a deep understanding of what it means to be human might make work teams more efficient. But the end result could amount to an assault on civilization.

THE QUESTION THAT WON’T GO AWAY

At the end of all this, one question hangs in the air, shivering from neglect: 

What does it mean to live well?

What does it mean to do good?

What is progress, and what is it actually for?

And what does any of that have to do with building and selling products?

Andreessen doesn’t seem to ask himself any of those. And that, in someone with his level of influence, is a dangerous omission.

Because the AI you design says something about the quality of attention you believe human beings deserve. The platform you build says something about the kind of community you think is possible. The system you fund says something about the kind of future you consider worth having. And if the person answering those questions has concluded that the inner self doesn’t exist, that introspection is neurosis, and that the great men of history didn’t stop to think too much...

…then the future being built has a very specific shape. And that shape doesn’t leave much room for you. For me. For any of us.

We don’t need the builders of the future to be philosophers instead of engineers. But we do need the engineers to also be human. To doubt. To examine themselves. To feel the weight of what they’re doing. To understand that “move fast and break things” is an easy line to throw around when the things being broken aren’t inside you.

And that real greatness, the kind worth having, is measured not only by the capacity to build empires. It is measured also by the capacity to elevate humanity while building them, and to keep it from sinking to the bottom of a poisoned Silicon Valley.