The Deep Feed

01 — Cal Newport

The Superintelligence Delusion

Why the race for a digital god is causing real-world wreckage

By Cal Newport · 8 min read

Editor's note: The obsession with existential risk is distracting us from the very real, very preventable harms being deployed today.

A strange fever has taken hold of the AI labs. Employees at companies like Anthropic and OpenAI are increasingly speaking of a future where their creations might wipe out billions of people. They talk about 'superintelligence' as if it were an inevitable deity waiting to be summoned. This rhetoric creates a massive media storm, a cycle of dread that feels heavy and significant. But there is a fundamental problem with this narrative: we have no idea how to build such a thing. The idea that we can simply scale Large Language Models (LLMs) until they spontaneously develop god-like reasoning is a convenient myth. It is a technical shortcut that allows researchers to bypass the hard, boring work of actual engineering in favour of discussing sci-fi scenarios.

The Cost of the Messiah Complex

When you convince yourself that you are in a race to save or destroy humanity, your ethical compass begins to spin wildly. The pursuit of superintelligence acts as a moral shield. If the goal is the ultimate prize, then the collateral damage of today feels insignificant. We saw this clearly with the recent HuggingFace hacking incidents. OpenAI took standard LLM-powered agents—tools used safely by millions—and stripped away their guardrails to see if they could perform cyberattacks. They weren't doing this because they needed to; they were doing it because they were desperate to hit a benchmark that would prove they were closer to a digital god. It was reckless, negligent, and driven by an ideology rather than a product roadmap.

The pursuit of superintelligence is a rhetorical crutch that lets you discuss sci-fi scenarios without engaging with technical reality.

This behavior is essentially the equivalent of putting experimental self-driving cars on a motorway to see if one can win a race. The risk isn't a sudden explosion of sentient machines; the risk is the cyber mayhem, the broken systems, and the erosion of digital safety that occurs when labs prioritise milestones over stability. We are being sold a future of extinction to distract us from a present of instability. The harm isn't coming from a super-intelligent mind; it is coming from the irresponsible hands of humans who are too focused on the horizon to look at the ground beneath their feet.

The real risks of the current AI trajectory:

Negligent experimentation with agentic tools
Prioritising benchmarks over safety protocols
The distraction of existential dread from practical security
Ideological rigidity in research directions

We do not have to accept this version of progress. AI development does not require a descent into chaos. If labs focused on producing useful, reliable tools rather than chasing a messiah, the industry could advance without the constant accompanying dread. The danger isn't the technology itself; it is the specific, reckless way we are trying to force it into existence. We are trading safety for a fantasy, and the price is being paid in real-world digital security.

Key Takeaway

The fear of a future AI god is a distraction from the very real, very human negligence happening in labs right now.

02 — Aeon

The Polymath's Advantage

Recovering the breadth of thought from the Islamic Golden Age

By Mariam Sabri · 10 min read

Editor's note: In an era of hyper-specialisation, the ability to move between disciplines is becoming a lost, but necessary, superpower.

Modern education is increasingly a factory for specialists. Students enter university with a narrow, vocational mindset, asking only one question: 'Will this make me employable?' This drive for immediate utility has created a fractured way of knowing. We have siloed mathematics from philosophy, and medicine from poetry. But the history of human progress suggests that the most significant breakthroughs did not come from people who stayed within their lanes. They came from polymaths—individuals who moved freely between disparate worlds of knowledge, testing the assumptions of one discipline against the realities of another.

Lessons from the Golden Age

The great Islamic polymaths—al-Biruni, Alhazen, Avicenna, and al-Khwarizmi—did not view knowledge as a series of separate boxes. For them, polymathy was a way of life. Take al-Biruni, for instance. He grew up in a state of scarcity, yet this lack of resources became his greatest teacher. Because he could not simply buy expensive scientific instruments, he taught himself to build them. This practical, hands-on understanding of how tools work allowed him to spot errors in the work of much wealthier contemporaries. He saw that a massive, expensive sextant used by a leading astronomer was actually sagging, distorting the data. His resourcefulness, born of necessity, gave him a clarity that wealth had obscured.

Breadth is not a luxury for stable times; it is the engine of enduring breakthroughs.

This kind of expansiveness offers a way out of the narrowness of modern thought. When we specialise too deeply, we lose the ability to see the connections that drive innovation. A mathematician might see a pattern, but a philosopher might ask if the pattern matters, and a poet might find the language to describe its impact. The polymathic tradition is about this very integration. It is a 'bricolage' of many worlds. By reintegrating these disciplines, we don't just learn more; we learn how to think more effectively about the complex, interconnected problems of our own time, from climate change to economic instability.

The core traits of the polymathic approach:

Cross-disciplinary testing of assumptions
Resourcefulness in the face of scarcity
Integration of theoretical and practical knowledge
A refusal to accept siloed thinking

We should stop fetishising 'disruption' through entrepreneurship courses and start teaching the art of breadth. The goal of learning should not just be to acquire a skill, but to build a mental toolkit that can adapt to any environment. The polymaths of the past did not just solve the problems of their age; they built the foundations for the centuries that followed. If we want to solve the problems of ours, we need to stop narrowing our vision and start expanding it.

Key Takeaway

True innovation requires the ability to bridge disciplines, not just master a single niche.

03 — Aeon

The Humoral Mind

What medieval medicine teaches us about emotional regulation

By Katherine Harvey · 9 min read

Editor's note: The distinction between 'demonic possession' and 'bodily imbalance' reveals a surprisingly modern understanding of mental health.

When we look back at the 12th century, we often see a world defined by superstition and cruelty. We imagine a man writhing on the floor, tied to stakes, being 'cured' by an exorcist because he was believed to be possessed by a demon. To a modern observer, this looks like pure ignorance. It is easy to dismiss medieval attitudes to mental health as nothing more than religious hysteria. But if we look past the hagiographies and the miracle stories, a more sophisticated reality emerges. Beneath the surface of religious explanation lay a medical tradition that understood the connection between the mind and the body with startling clarity.

The Balance of the Non-Naturals

The dominant medical theory of the time was the humoral system. Health was seen as a state of balance between four bodily fluids: blood, phlegm, black bile, and yellow bile. To maintain this balance, physicians focused on the 'six non-naturals'—external factors that directly impacted the human body. These included diet, sleep, exercise, and, most importantly, emotions. Medieval medical writers were not just talking about physical health; they were preaching a form of emotional regulation. Physicians like Juan de Aviñón warned that anger, irritation, and worry were not just unpleasant feelings, but direct threats to physical survival.

Emotional dysregulation was seen as a direct cause of physical illness, not just a symptom of it.

This wasn't just abstract theory; it was applied to real-world ailments. People in the 14th century explicitly linked their physical illnesses to mental stress. A merchant might suffer a fever after a stressful election; a daughter might suffer a fit of mania after a sudden fright. There was a recognition that intense emotions like fear and anger could physically alter the body—causing the blood to 'freeze' or the body to be 'consumed' by heat. While their terminology was different, their observation of the mind-body connection was remarkably similar to our modern understanding of psychosomatic illness.

Medieval observations on mental-physical links:

Anger as a heat-inducing, body-consuming force
Fear as a cause of sudden physical collapse
Stress and responsibility as risks for high-status individuals
Diet and sleep as tools for emotional stability

Studying these medieval perspectives helps us move away from the idea that we have 'solved' mental health through science. It reminds us that the struggle to manage the internal world is an ancient, human constant. The medieval physician's focus on lifestyle and emotional control as the foundation of health is a lesson we would do well to revisit. We have better tools now, but the core challenge remains: keeping the internal environment in balance to prevent the external world from breaking us.

Key Takeaway

The medieval understanding of the mind-body connection was less about demons and more about the practical management of emotional health.

04 — Not Boring

The Counter-Positioning Playbook

How startups use strategic misalignment to topple giants

By Packy McCormick · 7 min read

Editor's note: For a startup, the best way to fight an incumbent isn't to play their game better, but to play a game they cannot afford to join.

Most startups fail because they try to compete head-on with incumbents. They try to build a slightly better product, a slightly faster service, or a slightly cheaper version of what already exists. This is a losing battle. The incumbent has more capital, more people, and more brand recognition. If you try to win by being 'better', the incumbent will simply copy your features, drop their prices, and use their massive scale to crush you. To win, you don't need better technology; you need a different business model. You need counter-positioning.

The Trap of Incumbent Incentives

Counter-positioning is the practice of developing a business model that creates a conflict of interest for the incumbent. It makes it impossible for them to compete with you without destroying their own existing business. Take the example of Ramp. Traditional corporate card companies made money by encouraging customers to spend more—more spend meant more points and more fees. Ramp, however, built its business on helping customers spend *less*. If the big players tried to match Ramp's value proposition, they would be actively sabotaging their own revenue streams. Their shareholders would never allow it. This is how a startup buys time; it creates a zone where the giant is paralyzed by its own success.

Counter-positioning is the trickster power: it's how you attack a king without playing his game.

This strategy works because it exploits the very advantages that make incumbents strong. Their scale, their established processes, and their predictable revenue models become their shackles. A company like Base Power Company doesn't just sell batteries; they sell electricity through a model that makes ownership unnecessary. For a massive utility company to compete, they would have to fundamentally dismantle their entire infrastructure and business logic. They can't just 'tweak' their way into competing with a model that renders their core product obsolete.

Why counter-positioning works:

It targets the incumbent's conflicting incentives
It prevents direct price wars from being effective
It forces the incumbent into a 'lose-lose' decision
It buys the startup time to build deeper moats

If you are going to come at the king, do not attempt to out-resource him. Do not attempt to out-market him. Instead, find the part of his business that he is most afraid to touch. Find the part of his revenue model that he cannot sacrifice. That is where your opportunity lies. Counter-positioning isn't about being better; it's about being fundamentally different in a way that makes the incumbent's strength their greatest weakness.

Key Takeaway

The most effective way to defeat a giant is to build a business that they cannot copy without committing suicide.

05 — Not Boring

The Math of Machines

OpenAI's breakthrough in fluid dynamics and the future of automated reasoning

By Packy McCormick · 6 min read

Editor's note: When AI moves from generating text to solving Millennium Prize problems, the conversation about 'utility' changes forever.

The debate over whether AI is a threat to humanity often ignores the more immediate, staggering reality of what these models are actually doing: solving the unsolvable. Recently, OpenAI announced that a group of roughly 10,000 AI agents had produced a proposed solution to the Navier–Stokes existence and smoothness problem. This isn't just a clever bit of coding; it is one of the seven Millennium Prize Problems, a fundamental question in mathematics regarding how fluids behave in three-dimensional space. The agents worked for 88 hours using an unreleased model, producing a 166-page proof that was then formalised in Lean, a math programming language.

From Chatbots to Theorem Provers

This marks a shift in the capability of Large Language Models. We are moving past the era of the 'stochastic parrot'—models that simply predict the next likely word—into an era of automated reasoning. When an AI can construct a formal mathematical proof, it is no longer just mimicking human patterns; it is navigating the logical structures of reality. The controversy surrounding this breakthrough—with Anthropic accusing OpenAI of using their conversation logs—only serves to highlight how high the stakes have become. It is no longer about who has the best chatbot, but who has the most capable reasoning engine.

We are entering an era where AI does the heavy lifting of discovery, leaving humans to argue over the credit.

The implications for science are immense. If AI can tackle the Navier-Stokes problem, what does it mean for biology, materials science, or physics? We are seeing the beginning of a massive acceleration in the rate of discovery. Google DeepMind's AlphaGenome Atlas is another example, providing a petabyte-scale database of predicted molecular effects across the human genome. It doesn't replace experiments, but it tells researchers exactly where to spend their limited time and money. The AI isn't just a tool; it's becoming a navigator for the vast, complex data of the natural world.

The new frontier of AI capabilities:

Automated mathematical theorem proving
Large-scale genomic prediction
Complex document parsing and structuring
Accelerated scientific hypothesis testing

This is the real story of the AI revolution. It isn't about a sudden takeover by sentient machines; it's about the systematic automation of the most difficult cognitive tasks in human history. As these models move from generating content to generating truth, the boundary between human intellect and machine computation will continue to blur. The question is no longer whether AI can help us, but how we will manage a world where the hardest problems are solved by machines in a weekend.

Key Takeaway

AI is transitioning from a tool for mimicry to a tool for fundamental scientific discovery.

06 — Psyche

The Anatomy of Impatience

Reframing a personality trait as a manageable emotion

By Amy Arthur · 4 min read

Editor's note: Stop blaming your character for your frustration. Impatience is a feeling, and feelings can be managed.

Most of us treat our flaws as permanent fixtures of our identity. We say things like, 'I'm just an impatient person' or 'I've always been anxious.' By framing these tendencies as personality traits, we essentially surrender to them. We treat them as unchangeable weather patterns in our minds. But this perspective is fundamentally flawed. Recent psychological research suggests that impatience is not a fixed facet of who we are. It is an emotion—a temporary state that arises in response to specific circumstances. This distinction is not just semantic; it is the difference between being a victim of your temperament and being a manager of your emotions.

The Fairness Factor

Research from the University of California, Riverside, has shown that the intensity of impatience is often less about the actual length of a wait and more about our perception of fairness. We can endure a long delay if we believe the process is just and the wait is necessary. However, a short delay that feels arbitrary or unfair can trigger a disproportionate surge of frustration. This suggests that impatience is a reaction to a perceived violation of order or respect, rather than a simple inability to wait. When we understand this, we can begin to address the root cause of the feeling rather than just fighting the clock.

Impatience is an emotion to be regulated, not a character flaw to be endured.

The key to managing impatience lies in acknowledgement. Much like the character of Penelope in the Odyssey, who had to navigate decades of waiting, we must allow ourselves to feel the frustration without letting it define us. When we label the feeling—'I am feeling impatient right now because this delay feels unfair'—we create a small amount of psychological distance. That distance is where agency lives. It allows us to move from a reactive state to a reflective one, where we can choose how to respond to the delay rather than simply being consumed by it.

How to manage the impulse of impatience:

Label the emotion as a temporary state
Identify if the frustration stems from a sense of unfairness
Create psychological distance through observation
Focus on what can be controlled rather than the wait itself

We should stop apologizing for our impatience as if it were a broken part of our soul. Instead, we should treat it as any other emotion—something to be observed, understood, and regulated. By shifting our view from 'who I am' to 'what I am feeling', we regain the power to navigate the inevitable delays of life with more grace and less internal friction. The wait may be out of our control, but our relationship to it is entirely up to us.

Key Takeaway

Impatience is a temporary emotional response to perceived unfairness, not a permanent defect in your character.