Entry Shift

This story about how the New Work Foundation came to exist was a good read. It was founded by a former Meta and Salesforce executive as a nonprofit focused on Gen Z. She aims to solve for the entry lebel job disruption caused by AI and is among other senior tech leaders who stepping out of direct product roles to address the downstream effects of the systems they helped scale. Her vantage point inside companies already deploying AI agents at scale shaped a very consequential realization. AI is no longer just augmenting work, it is actively replacing large categories of entry-level and operational labor.

At Salesforce and Meta, she saw how quickly AI agents began absorbing tasks once reserved for junior employees: customer support workflows, basic analysis, content drafting, and coordination-heavy work. These are precisely the roles that have historically served as the entry point into major firms. As those layers compress, the traditional “first job” ladder is thinning, and the distance between education and meaningful employment is widening.

Her nonprofit response treats this not as a temporary labour cycle, but as a structural transition. If companies can scale output without scaling headcount, then Gen Z is entering a job market where learning-by-doing inside large firms is no longer guaranteed. The focus shifts from preparing for specific job titles to building adaptability in environments where roles themselves are unstable and frequently redefined by AI systems.

This is a model worth emulating across industries facing the same shrinking of opportunity of entry level professionals. Instead of assuming individuals will independently navigate disruption, it builds infrastructure for adaptation by helping young workers understand AI-augmented environments, identify transferable skills, and redesign career paths that do not rely on disappearing entry points. It is the much-meeded shift from participating in acceleration to building the tools that help others keep pace with it.

Career Fork

There seems to have been an increasingly frequency of posts in my LinkedIn feed from women in marketing who are taking off from the profession entirely. The new mothers make a good chunk of this population but it could also be my peers in marketing. There is a fork in the road in many marketing careers that becomes visible around motherhood. This has been true for the longest time but I do see some acceleration lately. Marketing is a predominantly female profession, particularly in executional roles like content, campaign management, and brand coordination. At the same time, it is a field defined by constant change with platform shifts, algorithm updates, and now rapid AI adoption that is reshaping how core work gets done.

Motherhood introduces a new kind of constraint: time discontinuity. The early months of a child’s life are intensely bounded and non-replicable, and for many professionals this creates a sharp reordering of priorities. Data consistently shows that mothers face measurable career penalties after childbirth, including slower progression, reduced access to roles, and higher rates of career interruption, even when they remain in the workforce.

At the same time, marketing itself is becoming less stable in its job architecture. AI is compressing tasks that once formed the backbone of entry and mid-level roles like drafting copy, producing variants, reporting, and basic campaign execution. Teams are becoming leaner, and expectations for output per person are rising. For someone stepping away temporarily, the question is no longer just about returning to the same job, but returning to a role that may have materially changed in structure and scope. 

In that context, the “clean break” becomes a rational response to overlapping uncertainty. It is not simply a withdrawal from work, but a choice between reinvesting in a fast-changing system after interruption or redirecting energy toward domains that feel more controllable and stable. The fork is shaped less by individual ambition than by timing: when both caregiving time and professional identity are under pressure from change, some choose continuity, and others choose exit. Women in my network it seems are choosing the later and I believe it may ultimately be the smarter choice to push reset.

Virtual Violin

I don't know the first thing about violins or violin-making but this Ars story about a CAD for violin makers caught my eye; listening to María Dueñas a few years ago was an unforgettable experience
The craft is centuries old and has lived in a tension between precision and mysterious art. Luthiers can measure wood density, carve to fractions of a millimeter, and replicate historical designs down to the last curve and yet, two instruments that look nearly identical can sound radically different once they’re strung and played. The difference is why violin-making is treated as part science, part inherited intuition, and part something closer to craft folklore than engineering.

The new tool developed at MIT tries to move that uncertainty earlier in the process. Instead of waiting until a violin is fully built to understand how it will sound, researchers have created a physics-based “virtual violin” that simulates how sound emerges from the instrument before any wood is carved. It starts with high-resolution CT scans of a real Stradivarius model and reconstructs the violin as a 3D system of millions of tiny elements. Each piece of wood, each string, and even the surrounding air is treated as part of a coupled physical system governed by acoustic and mechanical equations.

The result is not a sampled or AI-generated imitation of violin sound, but a first-principles simulation of how vibration becomes music. In the current version, the model can reproduce plucked-string sounds known as pizzicato and even render simple melodies by calculating how vibrations propagate through the instrument’s structure and into the air. Bowing, with its complex frictional interaction between bow hair and string, remains an unsolved layer of the problem, but the researchers see it as a next step rather than a dead end.

Of course, the obvious skepticism from craftsmen and musicians is whether this actually captures what matters. Violins are famously sensitive to irregularities: grain structure, microscopic variations in density, aging wood, tiny asymmetries introduced during hand carving. These are precisely the things that are hardest to model and easiest to dismiss as radnom noise until you realize they may be central to the instrument’s character. Even the researchers acknowledge they are not trying to reproduce the magic of historic instruments like Stradivari’s, but to map the physical space in which that magic might exist.

Delayed Trajectories

Reading about Charles Correa and the life he and Monika built around continuity and practice brought back a memory I had not revisited in years. In the early years of my career, I knew someone whose work stood completely apart from her peers. She had been offered an opportunity to work closely with a leading architect, the kind of opening that can shape an entire career. Taking that job would mean that she live and work independently in a different city. Her family insisted on different order of operations ostensibly out of concern for her safety. Marriage had to come first and then everything else would follow. She could not take that once in a lifetime opportunity.

At the time, it felt like a delay. But careers in fields like architecture don’t respond well to interruption as we saw in her case. The early years are when you build taste, networks, and a body of work that leads to the next commission, the next collaboration, the next idea worth pursuing. Miss that window, and the loss become non-linear because that one missed apprenticeship cascases to ten missed projects. For her that one deferred move turned into a life that never quite regained its original trajectory.

Reading about Correa got me thinking about the scale of his work and its continuity, the ability to return to ideas over decades and let them evolve. That kind of practice depends on uninterrupted time and self-direction. When those conditions are unevenly distributed, talent alone cannot compensate. We often talk about opportunity as if it were a discrete event, something you either take or don’t. But opportunity is also timing, and timing is shaped by who is allowed to begin, and when. Our friend was not allowed to begin.

Care Collapse

This  Atlantic reporting on childcare funding pressures tied to policy shifts under the current administration highlights a structural problem that predates any single administration but is now becoming harder to ignore. The U.S. childcare system has long operated on thin margins, relying on a fragile mix of parent fees, limited public subsidies, and underpaid labor. When temporary pandemic-era support expired, many providers were left exposed. The result is that centers are closing, waitlists are growing, and costs for families are rising at the same time. A co-worker told me recently that it would take her a year on the waitlist of the daycare center to get her newborn in. It was unclear what she would do until then as she and her spouse do not have the luxury of taking extended time off or leaving the workforce.

Data from multiple states shows thousands of providers at risk of shutting down or reducing capacity. Childcare is unusually sensitive to funding gaps because labor accounts for the majority of costs, and wages are already low relative to the skill required. Providers can’t easily cut expenses without cutting staff, and cutting staff directly reduces the number of children they can serve. That creates a feedback loop: fewer slots lead to higher prices, which further suppresses demand among lower- and middle-income families.

When childcare becomes scarce or unaffordable, workforce participation among women declines. This is not news and we've seen this happen for decades. Labor force data during and after the pandemic showed a clear correlation between childcare availability and employment rates. Businesses feel this through reduced labor supply, and governments feel it through lower tax revenue and higher dependency on social programs. In that sense, childcare is not just a social service but a core piece of economic infrastructure.

The policy debate, however, often treats childcare as a discretionary benefit rather than a foundational system. Short-term relief packages can temporarily stabilize the sector, but without a durable funding model. Expanded subsidies, public provision, or incentives for employers do not solve the problem fundamentally as the system remains prone to repeated shocks. What the current wave of closures makes clear is that the status quo is not stable. My former boss used to say in the context of problems wher no systemic solution is imminent that we are the cavalary. 

Executive Drift

Interesting Verge interview with the CEO of Uber that seems to make the case that the CEO role needs to morph into that of an AI orchestrator.

The conversation around AI replacing jobs has typically focused on routine or technical roles, but the more useful is if the CEO role itself becoming vulnerable. This discussion featuring Dara Khosrowshahi on Uber’s evolving use of AI, the focus wasn’t on eliminating leadership, but on how much of what leaders do is increasingly automatable. Coordination, optimization, forecasting, and even parts of strategic planning are now being augmented or in some cases handled entirely by AI systems.

Historically, CEOs have functioned as integrators: aligning information across teams, making decisions under uncertainty, and setting direction. But AI systems are rapidly improving at ingesting vast datasets, identifying patterns, and recommending actions in real time. In operational contexts like pricing, logistics, and marketplace balancing the areas core to companies like Uber, AI is already outperforming human-only decision loops. This raises a structural question: if large parts of coordination and optimization can be automated, what remains uniquely human at the top and I guess there is the matter of asking who is ideally suited to be in the position and how much should they be compenstated.

The conversation is about a shift not disappearance. The CEO role becomes less about processing information and more about framing decisions, setting constraints, and absorbing accountability. This is much like the AI ethics leader who need to take the fall as the AI cannot do jail-time. AI can suggest optimal moves, but it cannot own consequences in the same way a human leader must in areas involving regulation, ethics, and public trust. In that sense, the job may narrow but also intensify, concentrating around judgment rather than execution. Reading this got me wondering if the role should be handled by a revolving cast of characters instead of one that will create better balance strengths in judgment and subject matter expertise.

Lost Homes

I read an article recently about the melancholy of visiting a childhood home. The author talked about old drumsticks in a drawer, prom pictures on a bedroom wall, and the smell of a living room on a hot day. She described a physical ache at the thought of her mother selling the house, as if losing the building would mean losing a limb or a piece of her identity. I'd love to be able to relate but just can't.

Having moved seven times before I turned eighteen, there was no specific house that comes to mind when I think of a childhood home. It depends on the time window associated with some of the more memorable events of that period. If I tried to visit my childhood home, I would have track down a dozen different landlords who have long since painted over any trace of my existence.

The article mentions a study where a third of people travel back to their childhood homes as adults. They knock on doors and ask strangers to let them in. They cry in front of psychologists because the house is gone or changed. They speak of a narrative identity, a way to stitch their past self to their present self through four walls and a roof.

But what happens to your narrative when the setting changes every two years? When you are a military brat or the child of restless parents, your identity is not built on this kind of stable scaffolding. You learn early on that memories of good times with people are the only things you can keep; you learn how to say goodbye to them without looking back because you have to train yourself to look at every move like a new adventure to be able to cope with all that it brings.

There is a certain kind of privilege in being able to regress. The author talks about going home and letting her mom do her laundry or acting like a bratty teenager because the space triggers those memories. For those of us who moved constantly, there is no space to trigger the regression. We had to be adaptable. We had to be the new kid over and over again. Psychologists say that episodic memory is tied to location. They say our brains reboot to old activity patterns when we stand in the place where things happened. If that is true, my memories are scattered like widely disbursed with no central hub. There is no shrine to my younger self.

Maybe there is a different kind of wholeness found in the movement. While others are searching for a specific street corner to prove they were once young, I carry my history in the way I can make any anonymous apartment feel like mine within twenty four hours. This I've had plenty of pratice with.

Learning Facts

I first learned about Alpha school a few months back listening to an interview with the founder. Needless to say, her claims sounded a little too good to be true but I wanted to be open-minded. This is a brave new world for K-12 education and the current systems are falling short. Maybe it is time to think very differently, break the mold. Reading this Wired story gave me the parent's perspective which that interview missed.

In Brownsville, Texas, families once hopeful about Alpha School’s promise of cutting-edge, AI-driven education are now sharing deeply unsettling experiences. Marketed as a revolution in learning, where students achieve mastery in just two hours a day using adaptive software, Alpha replaced traditional teachers with “guides” who supervised but didn’t instruct. Parents like Kristine Barrios say the results were devastating: her 9-year-old daughter was trapped in endless math drills, denied snacks for missing targets, and ultimately developed anxiety, weight loss, and a deep aversion to learning. Alpha, which later lost access to its learning software for misuse, denied all allegations of mistreatment. That sounds like education dystopia and far worse than what we have going on in public school.

Alpha began as an elite experiment among Austin’s tech elite. Founded by MacKenzie Price and funded by automation billionaire Joe Liemandt, it was hailed as the future of education. It was supposed to be efficient, scalable, and data-driven. Tech luminaries like Bill Ackman and Reid Hoffman amplified its message, while the Trump administration and U.S. education officials praised its innovation. With backing from Silicon Valley and hedge fund circles, Alpha rapidly expanded across the country, positioning itself as a solution to the teacher shortage and a model for “AI-powered” schools. Truth be told with the right backers and AI in the name, it hardly takes genius to garner initial success these days.

Brownsville’s predominantly Hispanic, working-class community became the proving ground for Alpha’s 2 Hour Learning model. Lured by scholarships and promises of individualized progress, parents enrolled their children hoping for opportunity. The school’s design was seductive: cozy, modern classrooms, beanbags instead of desks, and rewards in the form of “Alphas,” a school currency redeemable for toys or trips. But behind the bright colors and motivational slogans was an unrelenting focus on metrics. Students worked silently under surveillance tools that tracked eye movements, screen time, and even behavior at home through webcams. This terrible metric collection scheme was never mentioned in the interview I had heard a while back.

What Alpha called “rigor,” many parents and former staff called cruelty. Students were punished for falling behind app-driven goals, and learning became synonymous with performance metrics. “Guides” with no teaching background and “academic coaches” hired overseas monitored student progress but offered little emotional or intellectual support. Critics say Alpha’s model ignored the improvisational, human side of teaching, the ability to sense frustration, adapt explanations, and nurture curiosity. As one parent put it, “They set her up for failure, and then punished her for failing. I seem to recall that the founder said the "guides" were in the classrooms bonding with kids, helping them find what energizes and motivates them so they could find their individual and unique path to success. It made me wonder if there are low-cost vs high-cost versions of the program. 

The school’s expansion plans ran into skepticism from education boards. Charter applications in Pennsylvania, Arkansas, and Utah were rejected as “untested” and lacking core subjects like history and social studies. Parents withdrawing their children found alarming gaps in basic literacy and writing skills, students who could read fast but not comprehend, or write without knowing how to form letters correctly. Some, like Barrios, spent months helping their children recover emotionally and academically from what they described as an experiment gone wrong. That is real tragedy for the kids involved in what seems to be a reckless experiment.

Alpha School’s story is a microcosm of a larger tension: the tech world’s conviction that software can replace teachers and standardize success. In practice, it exposed the dangers of reducing children’s education to data points and incentives. For the families of Brownsville, Alpha was not the future of learning, it was a warning about what happens when innovation forgets humanity. As one parent reflected, “All she needed was someone who could walk her through content that’s just not landing.” The dream of AI tutors may live on, but its human cost is now impossible to ignore. I wish interviewers would do a better job of teasing out fact from fiction, marketing and hyperbole.

Brain Drain

I've been following the story of Charles Lieber and see it as the proverbial canary in the coalmine. He was once the most cited chemist of his generation, former chair of Harvard's chemistry department, a man who brought in over $15 million in federal research grants and was widely considered a Nobel Prize contender, is now running a state-funded brain-computer interface laboratory in China. There might be a satisfying spy-thriller angle to his story but it seems like Charles Lieber didn't defect because China recruited him. He defected because America failed to keep him. More likely than not, the same system that failed him is the same system that is quietly failing hundreds of researchers whose names we don't yet know.

Colleagues who knew Lieber have described him as a man of almost pathological ambition, someone who worked six days a week, grew the largest pumpkin in Massachusetts because if he was going to grow a pumpkin it was going to be the largest one, and who had, by his mid-50s, done nearly everything a scientist in America could do except win the one prize that mattered most to him. The Nobel. People who knew him well believe the China relationship was, at its core, a Nobel strategy. A larger network, more graduate students, more prolific output, more international prestige. The $50,000 a month the Wuhan University of Technology paid him in cash, sometimes in brown paper packages, was almost beside the point for a Harvard professor of his standing. China was offering him scale, speed, and the infrastructure to do science at a level the American system had stopped being able to provide. He lied about it, repeatedly and on the record, and he paid a price. Two days in jail. Six months of home confinement and a $50,000 fine which is roughly one month's Chinese salary.

What Lieber found in Shenzhen is what an entire generation of American researchers is discovering: that the gravitational center of serious science has begun to shift. His new laboratory, i-BRAIN, gives him access to dedicated nanofabrication equipment and large-scale primate research facilities that were effectively unavailable to him at Harvard  because the American research model has become structurally incapable of deploying capital at the speed and scale that frontier science now requires. The fragmented US system of competitive grants, regulatory approval processes, shared equipment, and institutional bureaucracy was designed for a world in which America had no serious rivals. That world is long gone. China named brain-computer interfaces a national priority in its 2026 five-year plan. The Shenzhen ecosystem surrounding Lieber's lab has been funded with billions of dollars of government money. The building his institute will eventually occupy covers 750,000 square meters and is still under construction. The sign at the entrance to the broader campus reads, with a bluntness that almost demands respect: "Innovate with the Party." America's equivalent message to its scientists, delivered through the NIH grant review process and the indirect cost rate negotiations and the replication crisis and the adjunct crisis, has been considerably less inspiring.

This would have been unthinkable twenty years ago. For most of the twentieth century and well into the twenty-first, the traffic in scientific talent moved in one direction. The best minds from everywhere in the world came to America. They came for the universities, the funding, the freedom, the meritocracy whether real or imagined, and the simple fact that America was where the best science happened. China was a place you left to do serious work, not a place you went to. The researchers who returned to China from American institutions were going home, making a sacrifice, trading opportunity for family or patriotism. Lieber is something categorically different: a native-born American, a Harvard institution unto himself, who looked at what China was building and concluded it was the better bet. 

Lieber was never charged with transferring a single piece of intellectual property to China. He was convicted of lying about his involvement, which he did, on video, in terms that made his defense attorney's job nearly impossible. The prosecution made enough noise to damage American science's relationship with Chinese researchers for years but it did not stop Charles Lieber. He served his sentence, waited out his house arrest, and is now in Shenzhen doing the work he always wanted to do, with better equipment, more funding, and the full backing of a state that has decided his field is a national priority. A former NSA general counsel called him Exhibit A in the failure of American legal deterrence. He is also, less comfortably, Exhibit A in something else: what happens when a country decides that winning a geopolitical competition is more important than actually building the conditions in which its best people want to stay.

Zombie Capitalism

Deep in the tropical forests of Southeast Asia, a carpenter ant climbs a plant stem, bites down on a leaf vein at a precise height, locks its jaw, and dies. It didn't choose any of this. A parasitic fungus called Ophiocordyceps unilateralis hijacked its muscle tissue but not its brain and walked the ant through a programmed sequence of behaviors designed entirely for the fungus's benefit. The ant's body moved. The ant was gone. What makes this particularly unsettling to scientists is that the fungus never even touched the brain. It didn't need to. Control the muscles, and the mind becomes irrelevant.

This disturbing image is also, increasingly, a useful metaphor for the modern workplace.

In today's job market, the power dynamic between employer and employee has shifted dramatically and AI is providing the perfect cover story. Algorithmic scheduling systems tell workers exactly when to show up, how long to spend on each task, and when their "productivity score" has dipped below acceptable levels. Warehouse workers follow routes dictated by an app, with deviation tracked in real time. Call center employees are scored by AI on tone, pace, and word choice. Gig workers are dispatched, rated, and quietly deactivated by systems they never see and cannot appeal to. Like the fungus, these systems don't need to control your mind. They just need to control your movements and the behavior follows.

This  mirrors the biology quite closely in that the mechanism bypasses conscious resistance. The ant doesn't fight back because it doesn't experience the fungus as an external force. It simply moves. In the same way, many workers don't experience algorithmic management as coercion. They experience it as "how the job works." and if they are junior enough, the only way they know how the job works. The AI sets the pace. The AI flags underperformance. The AI suggests when you should take a break. And because it's framed as neutral, data-driven optimization rather than a boss cracking a whip, it's far harder to push back against. You can argue with a manager but can't with an algorithm.

When Ophiocordyceps finishes with its host, the ant's body is simply the vehicle through which the fungus spreads, the ant bears the cost, the fungus reaps the benefit. In the modern workplace, AI is filling the same structural role. When a worker is let go because "the data indicated performance issues," the employer isn't the villain, the algorithm is. When wages are kept low because "the market rate tool determined the band," there's no human decision to contest. When a shift is cut because "scheduling optimization identified low demand," the conversation is over before it starts. The fungus which in now the system has made the call. The employer is just the forest floor where everything plays out.

This outsourcing of accountability is not accidental. Companies have discovered that AI provides a layer of deniability that human managers never could. A bad boss is a liability. An AI system is infrastructure. And crucially, infrastructure has no feelings to hurt, no biases to prove, and no deposition to sit through. Workers across industries are reporting the same experience: decisions that once required a conversation now arrive as notifications. Appeals go nowhere because there is no one to appeal to. The ant, it turns out, doesn't need to know it's infected. It just needs to keep climbing.

The carpenter ant has no recourse. Once the fungus takes hold, the behavioral sequence runs to completion. The researchers at Penn State who uncovered how Ophiocordyceps works did so precisely because they asked the right question. They did not ask "what is the ant doing?" but "who is actually in control here?" That same question is worth asking in every workplace where AI has been handed the wheel. Whose interests does this system optimize for? Who benefits when I climb to the designated height and bite down on the designated leaf? And if I stopped, would anyone notice, or would the spores just find another ant?

The zombie ant fungus has been perfecting its system for millions of years. Algorithmic management is only a few decades old. The difference is that workers, unlike ants, can still name what's happening to them and naming it is always the first step toward something the ant never had: a choice.

Expat Surge

According to this Economist story, something significant is happening across the Western world, and it's getting far less attention than immigration debates. People are leaving their home countries in record numbers. Apparently, roughly 4 million people emigrated from 31 Western nations in 2024 which is about 20% more than just before the pandemic. The numbers are striking across the board: Swedish emigration is up more than 60% since 2019, New Zealand's is up 29%, and Canada's departures rose 34% in just the third quarter of 2025. Iceland recently recorded its highest emigration level ever, and Italy's statistics office described the trend as a full-blown "boom."

Three forces are driving the surge. First, the pandemic permanently normalized remote work, making it far easier for people to imagine living abroad while keeping their careers intact. American multinationals now employ 36% more people overseas than they did in 2019. Second, taxes are pushing high earners out, with top effective rates in both the UK and US near historic highs, making low-tax destinations increasingly attractive for those in peak earning years. Third, and perhaps most telling, is politics which is not surprising. Disillusionment is cutting across party lines. There's Americans fleeing Trump, Britons escaping what some call Starmer's socialist agenda, Canadians tired of over a decade of Liberal rule. Research backs this up: studies find a direct link between democratic decline and rising emigration. The politics of one's home country can feel intolerable because the person does not get what they voted for but much worse things abroad are impersonal as they have no claim or stake in it.

The consequences for the countries being left behind are real. When governments invest in educating young people who then emigrate, they lose future tax revenues. This is a particularly painful hit for smaller nations with aging populations. Research focused on central and eastern Europe found that emigrants tend to be more liberal than those who stay, and their departure has coincided with democratic backsliding in their home countries. The brain drain is not abstract: educated people in their 20s are disproportionately the ones leaving, with New Zealand data showing degree-holders at least twice as likely to emigrate.

That said, for every country losing talent, another is gaining it. The number of Americans in Germany has risen over 60% in the past decade, and the Netherlands has absorbed an outsized share of new Western arrivals. Many emigrants eventually return home, bringing savings, skills, and global networks with them. And those who don't often remain connected through diaspora communities. In an era defined by nationalist rhetoric and closing borders, the quiet, record-breaking movement of Western citizens is quietly making the world more interconnected than the headlines suggest.

Dosa Rising

I am far from an expert masala dosa maker but can do a decent job. Even at that level, the production has never failed to impress the unitiated. It's great to see it making it up the pancake lederboard.

The best masala dosa of my youth was made by a street vendor who had perfected both taste and execution. It was crispy, fermented, filled with spiced potato, and now offically one of the greatest pancakes on earth. TasteAtlas has ranked masala dosa sixth on its May 2026 list of the top 100 pancakes worldwide, awarding the dish a 4.3-star rating. The travel and food guide, which bases its rankings on a blend of user ratings and expert reviews, has essentially confirmed what millions of South Indians have known for generations. Once a South Indian breakfast staple, the masala dosa is now rubbing shoulders with the world's finest. Though, it had made its way to the rest of India long before being ranked among the world's best pancakes. I've had very decent masala dosa in Kolkata which would have been unthinkable and impossible when I was growing up.

The humble foundation of dosa has spawned an entire family of variations. They range from Mysore masala dosa and rava dosa to paper dosa and even an avocado dosa for the Gen Z crowd.  The last one I have some trouble accepting but then I am not the target. The dosa's crisp texture, tangy fermented flavor, and versatility have won fans well beyond India, and the TasteAtlas ranking makes that global reach official. Notably, the dosa family dominated the list: regular dosa came in at No. 15, while the ultra-thin paper dosa landed at No. 35. I've always liked the plain dosa better than all variants but most people in my family prefer the ultra-thin paper dosa. 

Indian chefs are taking the recognition as a cultural milestone, not just a culinary one. Chef Hiten Puri described the ranking as a clear sign of the rise of India's soft power, where food becomes a cultural ambassador. Chef Chetna Chopra, who introduced avocado dosa to Delhi, went further, calling the dosa "the most intelligent pancake" and noting that it represents one of India's oldest expressions of fermentation, something that was part of everyday cooking long before it became a global wellness trend. For her, the irony is that the world is only now discovering something Indians have eaten their whole lives.

The ranking also arrives at a moment when dosa has been making headlines for less flattering reasons. A recent social media post went viral after author Ana Huang wrote about dosa in her book King of Gluttony with a character ordering it with eggs, strawberries, and cake, a somewhat bizarre combination that sparked widespread mockery online, with one commenter joking they'd never seen those words in a sentence together before. It's a funny reminder that as Indian food gains global prestige, it still has some explaining to do. On the global leaderboard, the masala dosa sits behind favorites like Latvia's kartupeļu pankūkas, China's jianbing, and France's crêpes, but given the trajectory, it may not stay at sixth for long.

Smart Vibrations

As someone who has been hearing impaired for a very long time, I love reading about technology that solves problems for people like us, specially those whose impairment level is much worse than mine.

For deaf and hard-of-hearing people, a blaring smoke alarm or burglar siren in the middle of the night is essentially silent. Existing solutions like vibrating devices placed under the pillow, have helped, but they've come with a frustrating trade-off. While effective, many users find them bulky and uncomfortable, and nobody wants an uncomfortable pillow keeping them up night after night, so they often go unused. Researchers at Nottingham Trent University set out to solve that problem, and the result is a remarkably simple idea: a smart sleeve that slips right over the pillow you already love.

Rather than building electronics directly into a pillow, the team created a thin, flexible sleeve that fits over a standard pillow and inside a typical pillowcase, allowing users to keep their preferred bedding while gaining the benefits of an integrated alert system. The secret is in the fabric itself. The sleeve contains four haptic actuators each measuring just 3.4mm by 12.7mm  embedded directly into a yarn-like textile structure, making them completely undetectable while you sleep. But when triggered, those tiny components pack a punch: they can vibrate intensely enough to wake even the heaviest of sleepers.

What makes the device especially thoughtful is how it communicates. The electronics connect to a smartphone via a microcontroller, which links wirelessly to household alarms and uses different vibration pulses to distinguish between a fire alarm, a burglar alarm, and an incoming phone call. So a user isn't just jolted awake,  they're told why they've been woken up, which could make all the difference in an emergency. The project was born directly out of conversations with the Deaf community, whose feedback made it clear that comfort was just as important as function.

The invention is still in prototype phase, and the road to market has some hurdles. Research supervisor Dr. Theo Hughes-Riley noted that large-scale production poses a challenge, as the current manufacturing process involves quite a few steps done by hand. Still, the team has confirmed the sleeve holds up to real-world use. Most importantly, the yarn has passed rigorous durability tests including multiple washing cycles. With the right industrial partner, this could be one of those rare innovations that is both genuinely life-saving and genuinely comfortable; proof that good design and good intentions don't have to be at odds.

Six Months

We once believed human intelligence was a fortress, built on the slow accumulation of knowledge and the discipline to apply it. For generations, entering the middle class meant mastering the rules of that structure: learning to memorize, to repeat, to serve as a reliable cog in a private or public machine. But as machine intelligence expands past every border we assumed would hold, that old contract is evaporating. Near-infinite, zero-cost cognition has exposed an unsettling truth: what we spent lifetimes training our children to do, a machine now does for the asking.

An economic hourglass is taking shape. The traditional ladders of upward mobility are dissolving from the bottom up. At the top, a small number of people orchestrate vast, autonomous systems, and their wealth and influence multiply into something like a new aristocracy. At the bottom, the masses are settling into a fluid, fractional existence, tethered to algorithmic platforms that manage intellectual labor the way factories once managed physical assembly lines. For now the machine outpaces the old paths to security. Whether bankruptcy follows is still an open question.

Educational systems built for a world that no longer exists make this feel more ominous still. We keep sending young people into a workforce model that's disappearing and expecting them to find their own way to survive it. The tool for success is available to everyone. The cultural and psychological capacity to use it well is not. It rewards a rare mix: creative independence paired with real numeracy and literacy, the kind of foundation that mostly gets built in homes that are educated and financially secure. That entry requirement locks out most people before they've had a chance to try.

Human value is starting to be measured by clarity of intent. White-collar workers now have to ask what it means to contribute when the mechanical grunt work of thinking is free. I see it at work as an identity crisis: people don't know what to do with the time that's opened up, so they keep doing what they've always done, even as the absurdity of it stares back at them. A product sits ready to ship, and customers could give feedback directly without a bloated leadership tier standing between them, but the stakeholders won't relinquish control. They delay the release six months to align on what needs to be built. In that same six months, agile competitors without the legacy operating model take more than half the addressable market. By the time shipping becomes politically acceptable, there's no market left to ship into, and the goalposts for winning have moved somewhere the old plan can't reach anymore.

Saree Longing

I can’t recall how I first stumbled upon this Facebook page where a woman talks about handloom sarees, but it’s become one of my favorite places to stop after a long day, when I need something that takes my mind off things. I have no memory of buying a saree for myself since I left India. I received my fair share of gifts, of course, but none that I would consider a family heirloom. I love sarees, and if it were up to me, that’s what I would wear every time I stepped out of the house. There are hopes and dreams, and then there is reality, in which I wear a saree maybe once in a decade. Perhaps that’s why the dreams feel so oddly disproportionate.

This woman fills that aching void between who I want to be, in terms of how I dress, and the limitations of real life that dictate what I can actually wear on a regular basis. Listening to her explain how to spot a real Jamdani from a fake one feels deeply satisfying — as if knowing those details might somehow make up for the fact that I own neither real nor fake versions of any of the sarees she talks about. When I got married in my twenties, I received the de rigueur set of Banarasi, Dhakai, and Baluchari sarees from my parents and parents-in-law. The rest were handloom silks and cottons gifted by relatives. They were too numerous to bring to America, so I left them with my mother and never really had an occasion to get them back.

With separation and divorce in a foreign country came the loss of Indian friends who had known us both and were unwilling to choose one of us over the other. They chose instead to part ways entirely. The same fate befell many bonds with relatives who felt I had acted unwisely and in haste. While all of that is water under the bridge now, it meant I no longer had a place for sarees in my life — no use for an article of clothing that had been my most favored since I became an adult.

I wore sarees regularly from my teens onward, an oddity among peers who saw them as a proxy for the loss of freedom and youth. I never did. To me, the saree was always the most dramatic, versatile, and dynamic thing a woman could wear. It allowed her to express herself in infinite ways. To me, it was a garment of power and possibility. I mourn that loss to this day, which is perhaps why I love listening to someone speak with such obvious love and intimacy about it.

Passport Arbitrage

This Econmist piece takes the reader to the streets of Da Nang at night that are a neon fever dream, the humid air smelling of grilled meat and salt water. In the bars along the coast, men from the West sit with single beers, scanning the room intensely. They are the Passport bros and believe they have finally cracked a code. To the casual observer, they are merely tourists or remote workers. But to themselves, they are refugees from a domestic dating market they describe as terminally broken, seeking a life where the math of attraction (hopefully leading to cohabitation and family formation down the road) finally adds up in their favor.

The movement is often framed as a simple life hack. The logic simple: if you cannot find the partner you desire in a culture where your social and financial capital is average, you move to a culture where those same assets are extraordinary. In the United States, a data analyst living in a parent’s basement might struggle to secure a second date. In Vietnam or Thailand, that same individual becomes a pillar of stability and a gateway to a different life. It is a migration driven by the same evolutionary pressures that have always guided mate selection, yet it is being accelerated by technology that has turned dating into a high-stakes ranking system.

The philosophical roots stem from what come to be known as manosphere. There, the world is divided into those who are awake to the supposed injustices against men and those who remain oblivious. The Passport Bro takes what they call the passport pill, a belief that the modern Western woman has been conditioned to reject traditional femininity and masculine authority. By leaving, these men are seeking sun and lower rent along with a return to an imagined past where gender roles were fixed and the domestic hierarchy was unquestioned.

This yearning for a lost order is a recurring theme in history. Whenever social norms shift rapidly, a segment of the population inevitably looks backward with a sense of mourning. After the Second World War, thousands of American soldiers returned home with war brides, praising women from abroad for being less competitive and more accommodating than their domestic counterparts. Today, that same sentiment is repackaged for the era of the algorithm. Instead of war stories, there are TikToks ranking the most submissive nationalities and YouTube tutorials on how to find a woman who will take pride in making a man’s life easier.

The tragedy of the movement lies in its transactional nature. While many of these men speak of love and companionship, the language they use is often stripped of romance and replaced marketplace vocabulary. They talk of rosters, rankings, and hypergamy. They measure their success by how much more their salary and ethnicity count for in a developing economy. It is a system built on an imbalance of power, where the primary attraction is often the very inequality the men claim to be escaping. It seems to me that very mindset brought on the struggles in the local dating market for them. There is no telling what the imported bride will turn into after a few years in America where she will have choice, access, and freedom. To main the leverage that got them to Da Nang in the first instance, they will likely need to remain there which can be fraught with innumerable complications they don't see at first. 

There is a somber irony in the fact that many Passport Bros claim to be seeking something authentic and traditional while relying on the most modern of tools to find it. They use Google Translate to communicate with partners they cannot speak to and use social media to broadcast their curated lives to an audience of envious peers back home. They are trying to build a future out of the fragments of a past that likely never existed in the way they imagine.

The rise of the Passport Bro is a symptom of a much deeper fracturing in the West. Men and women are drifting apart, separated by different expectations of what a relationship should be and who should hold the power within it. As the cost of living makes the traditional breadwinner model nearly impossible in the United States, some men find it easier to move the goalposts than to change the game. They fly halfway around the world to find a place where they can finally feel like the winners they were told they were supposed to be, leaving behind a home they no longer understand for a paradise built on a temporary advantage.

Hard Things

Arthur Brooks’s piece on Schopenhauer and big, difficult projects came at a time when I've been thinking about a project. I've had the urge to do something hard not because it’s practical, profitable, or even visible to the world but becasuse it pushes the boundaries of what I believe I can get done. Admitedly the bar is not very high but life experience shows that incremental progress does add up.

Schopenhauer’s insistence on sustained effort feels almost unfashionable now. We’re encouraged to pivot, to iterate, to abandon what doesn’t immediately respond. Fail fast and break things along the way is how most of the tech business moves. The enduring value of an approach that's so inherently brittle is questionable. We live in a world where peristence is framed as stubbornness unless it pays off fast. But there’s something clarifying about staying with a thing long enough to let boredom seep in, long enough to notice the urge to quit not because it’s impossible, but because it’s demanding. That is when you get to make a real assessment of the value of what you are pursuing.

Big projects aren’t heroic every day. Most days, they’re boring. They ask you to show up when no one is clapping and progress is invisible, when quitting would be entirely reasonable. We are handed more than a few such projects by life but a self-chosen challenge is a different thing entirely. You suffer, but for a reason you consented to. That consent changes everything. It turns struggle into meaning instead of resentment.

Since reading this artcle, I've been thinking about why we’re drawn to these projects even when we don’t need them. Maybe it’s proof—to ourselves more than anyone else—that we can still concentrate, still persist, still finish something in an age designed to fracture our attention. Finishing it would change who we are and expand the range of what is possible.

Job Paradox

I've had experience working with companies that are all-in on AI and want every last person to skill up. There is ample support and opportunity to reinvent one's job, be recognized and rewarded for it. Sadly, their offshore teams in India were the most resistant to change and continued to follow their old ways of working unless forced to comply. There is a pervasive fear of job loss due to AI so folks refuse to use it and consequently do not develop the skills they need to survive it. So reading this story about how India will magically adapt and win with AI felt out of sync with reality on the ground.

The "AI job paradox" sounds like a nice solution to a messy problem. The theory is simple: as machines take over routine tasks, they free up humans to focus on high-level work—empathy, judgment, and complex strategy. It’s a comforting vision for India’s workforce, but it ignores several hard realities.
Automation doesn't just change tasks; it deletes them at scale. Much of the process drag in organizations are a result of needing many people cross functions to work in co-ordination to get something one. This is mostly a relic of the past. One person managing an AI tool can often produce what twenty workers used to. Even if "higher-value" roles appear, they won't appear in the millions. For a country that needs to create a million jobs every month just to keep pace with its population, a net-loss in total roles is a crisis, not a "shift."

Then there is the reskilling myth. The idea that a data-entry clerk can simply pivot into an "ethical judgment officer" is optimistic at best and delusional at worst. These high-value skills—like nuanced decision-making—are usually the result of elite education and many years of experience. We aren't looking at a workforce moving up; we are looking at a workforce being hollowed out. The top 5% will thrive as AI orchestrators, while the rest are pushed into low-wage manual labor that remains too cheap to automate.

The article claims humans are indispensable because AI makes mistakes. A lot of people cling to this idea very passionately and refuse to entertain any other possibilities. But history shows that companies often accept "good enough" results if they save money. We’ve already seen this in customer service and content moderation. Instead of being empowered "orchestrators," many workers will likely become "rubber-stampers," clicking "approve" on machine-generated work for lower pay. I've had tenured engineers tell me that catastrophe will follow if we use AI to build and mantain API inegerations for example. They have chosen not to understand why infact this is one for the ideal use cases for AI; the so called spec-driven autonomous loop. It would be nice to see it attempted instead of declaring code red at the very thought of it. 

I was chatting with an offshore engineering manager recently ans he shared how difficult its been for him to bring his team along because they code manually and do not want to let go of their ways of working because it effaces their professional identity. He is trying to motivate them by saying they have the choice of falling in line or being replaced with one senior dev in America who runs a team of agents the same size of the current team in India. The compensation math would work out. His team has understably pushed back saying what is to prevent that from happening already. These are real issues and there are no good answers. He has been able to bring half of his team along to the new ways of working so far. 

AI moves at the speed of light. Indian education and corporate training move at the speed of a glacier. By the time our training programs catch up to the "skills of the future," the technology will have moved again. The paradox is that the more "human" a job becomes, the more exclusive it becomes. Unless we address the massive disparity in training and the raw loss of volume, the AI wave won't lift all boats—it will just widen the gap between the shore and the ship.

Corporate Betrayal

The corporate world has long been tried to build a constructed facade of pseudo family where companies invest millions in culture and people developmen supposedly aimed and deepening emotional ties with their staff. During prosperous years these organizations promote a narrative of mutual loyalty and shared values and purpose using language that mirrors the bonds of a household to encourage extra labor and commitment. The insinuation is the employer is a partner who will provide support through thick and thin. However as economic conditions shift these organizations reveal their true colors by dismantling the very benefits that define their purported care for the collective.

The recent actions of Deloitte serve as a stark example of this shift in character during a period of perceived crisis. Despite reporting billions in revenue the firm moved to slash essential benefits for a specific tier of its workforce including halving parental leave and eliminating significant reimbursements for IVF and adoption. Much like a partner who suddenly withdraws financial and emotional support for a shared future when conditions get tough, this move signals exooses thd family rhetoric for the lie it always was.

This betrayal is compounded by the use of falsehoods to mask the harsh reality of these reductions. By framing the removal of vacation time and parental support as a modernization of talent architecture these companies engage in a form of professional gaslighting. This behavior does not square with any prior activities for culture development because it fundamentally undermines the trust required for a community to exist. When a partner rebrands a lack of support as a necessary evolution for your own good the relationship turns null and void.

Zoom offers another clear instance of a brand showing its true colors after building its identity on the concept of human connection and happiness. After becoming a household name by facilitating relationships during global isolation the company began scaling back the perks and cultural initiatives that it once touted as its competitive advantage. This reveals a hierarchy of worth where the employees who built the brand are treated as expendable line items once the initial growth phase slows. I hope the current tide turns in a way that big companies become a relic of the past.

Returning Cursive

The Futurism piece about "AI-native" college graduates arrived in my feed the same week I found J's elementary school workbook at the back of a closet shelf, the one with the ruled pages and the letters she'd traced in pencil, each one getting marginally better across the page. It felt like a relic of the past that I should preserve. By 2012 teachers across most of the country stopped teaching what was in it, because the Common Core standards adopted two years earlier had swapped cursive instruction for keyboarding, and by 2012-13 the substitution was effectively complete.

The Futurism piece quotes a New York financier, anonymous, telling the Financial Times that his firm now actively avoids AI-literate STEM graduates and hunts for humanities majors instead, on the theory that humanities students still have critical thinking. My first thought was this person will be deeply disappointed in due season. The research on what happens when children stop writing by hand and move to keyboards is not new, and it is fairly consistent: handwriting of any kind, in elementary grades, produces better letter recognition, better reading comprehension, and more widespread brain activation than typing, because tracing out letters by hand involves sensory, motor, and cognitive systems simultaneously in ways a keyboard doesn't. The Common Core didn't made acurricular and neurological choice, for roughly forty million children, with very little public discussion about what that meant downstream.

The humanities major that financier is betting on graduated in 2025 was in first grade in 2012. Depending on which state she grew up in, she may have had no handwriting instruction of any kind past second grade, and the AI she used to write her papers on Middlemarch is the same AI her computer science classmate used to write code. The intuition that a literature degree protects you from cognitive outsourcing is not wrong exactly, but it is about ten years behind the actual sequence of events. By the time ChatGPT arrived in 2022, the groundwork had been laid through a decade of telling children that thinking slowly, by hand, on paper, was a skill the future didn't need.

Twenty-five states now require cursive instruction again. California reinstated it in 2024 for grades one through six, which made headlines, possibly because California had been the one to drop it so visibly in 2010. The researchers who study this are careful to say it's handwriting in general, not cursive specifically, that shows cognitive benefits. What the kids need is to form letters with their hands, repeatedly, over time. Whether the letters are connected or not turns out to matter less than the act of forming them at all. ChatGPT and the like completed the job started by the Common Core decision over a decade ago.

Performative Empowerment

I am not sure how I feel about the "reverse baraat." The road to hell is paved with good intentions, and while meant as a radical act of love that rises above social stigma, using the specific rites of marriage to celebrate its dissolution is a misguided approach that oversimplifies complex trauma. A wedding is a sacred ritual signifying union; applying those same celebratory markers to a divorce—the collapse of that very unit—feels like a mockery of the original sanctity of the vows. As someone who was divorced long ago and felt the full force of social shunning in India, I am certain something like this would have ill-served me. Deploying public pomp and circumstance risks turning a painful family transition into a performative spectacle that prioritizes optics over the meaningful support a woman actually needs.

Notwithstanding the pretension of being modern and liberated, a ceremony like this consolidates the daughter’s status as a "ward" whose identity is defined by her relationship to her father. He gets to give her away or take her back at his discretion, making a huge show of both. Sadly, everything remains focused on the father’s role as the protector and the daughter as the recipient of his public approval—the last thing a divorced woman needs after such a life event.

There is also an undeniable emotional toll to such a "show." Divorce involves complex, confusing emotions that must be experienced and expiated. Forcing a festive atmosphere onto someone who has just endured a grueling legal battle creates an environment of toxic positivity, where the woman feels pressured to appear happy and empowered for the sake of her father’s public stand. While well-intentioned, the family may unintentionally be trapping her in the same cycle of performance and social approval she was trying to escape, rather than allowing her the space to redefine herself on her own terms.

Fragile Promises

Reading about the 401(K) millionaire being the new goal to achieve makes me wonder if there isn't a catch somewhere; if it wasn't written by AI to further its own cause. Lately it's been framed as the crowning achievement of the American middle class, yet current market data suggests this status is far more precarious than the headlines imply. A significant portion of this growth is not the result of broad economic health but is instead driven by extreme market concentration. Because a handful of technology giants responsible for the current artificial intelligence boom represent a disproportionate share of index gains, the average retirement portfolio is now dangerously sensitive to a single sector. If the projected return on investment for these technologies fails to materialize, the valuations supporting these million-dollar balances could correct sharply, revealing that much of this perceived wealth was merely a byproduct of a temporary speculative cycle.

Many retirement accounts that appear diversified actually hold fixed-income products and bond funds that are increasingly tied to the massive capital expenditures required for data centers and infrastructure. This creates a feedback loop where even the supposedly safe portions of a 401k are vulnerable to the same industrial shifts as the stock market. When the underlying debt is used to finance growth that has yet to prove its long term profitability, the stability of the entire retirement vehicle becomes a mathematical illusion rather than a guaranteed safety net. Such fawning over the 401(K) millionaire status, is a way to nudge people to put more into these accounts promising them the illusion of financial freedom which may not materialize in the end.

Since a traditional 401k consists of pre-tax dollars, a seven-figure balance is effectively a joint account with the government, with a large percentage of the total earmarked for future federal and state liabilities. When adjusted for modern inflation and the high cost of living in major urban centers, a million dollars no longer provides the permanent financial independence it once signified. Data shows that many individuals in this bracket still carry significant personal debt and high monthly overhead, meaning their wealth is often spoken for long before it is even withdrawn.

Saying No

Read a LinkedIn post recently where a PM coach broke down how to say no, four scenarios, four scripts, each carefully matched to who's asking. I have seen PMs do exactly as he prescribes and have very mixed feelings about this approach. The reason PMs are saying no constantly is that nobody has been clear enough about what the product is actually for.

I worked with a woman once who ran product at a mid-size logistics company. She could say no beautifully to anyone who came between her and her roadmap. Her product reviews were clean and defensible and features accumulated in the way a dorm room collects furniture from people who move out: nothing fit together, nothing was ever removed. What she was doing, very skillfully, was managing the symptom of not knowing clearly enough what the product was meant to do and specifically not meant to do. A scalpel manufacturer that adds a bottle opener because engineering freed up capacity hasn't gotten better and faster at losing the thread. This is not product management is about.

The AI question changes the terms here. When building something took six weeks and three engineers, friction enforced discipline. Now you can prototype in a day and run a feature behind a flag to see if it moves anything. The cost of yes has dropped considerably. Which means the PM still practicing "no" as a primary skill, with scripts calibrated to audience, is optimizing for a constraint that's loosening. If the thesis is sharp enough, most of the triage happens before the ask arrives. The VP's request doesn't need to be responded to with why it pulls the product away from what it's actually trying to be, not "it doesn't rank against our top three objectives," which is process language for the same lack of clarity.

The harder version of the job the post doesn't get to: what happens when the thesis itself is wrong. When the scalpel market is collapsing and the task isn't holding the line but making the case, clearly enough that the organization can actually move, that this is the wrong direction and here is what might be right instead. The post's version of hard is telling the VP no without damaging the relationship whereas the actual hard thing is being the person who says, when everyone has been pulling in the same direction for two years, that the direction was wrong, specially that the person is not the senior-most person in the room and yet it is their job to call these things out.

Digital Leviathan

Palantir's "The Apotheosis of Data,"is not your typical corporate whitepaper. Where most tech companies dress up their ambitions in the language of connection and community, Palantir has gone a different direction. The central thesis is: philosophical weight, civilizational stakes, and the unmistakable conviction that Western decline is, at its core, a data infrastructure problem.

CEO Alex Karp has never been shy about ideology, and the manifesto reflects that. It's part strategic vision, part cultural polemic, invoking concepts like "transcendence" and "essential destiny" in the same breath as software architecture. The argument, stripped of its rhetorical pretences, is that liberal democracies are losing ground because they've been slow to weaponize their own information advantages, and that Palantir exists to fix that. It's a serious claim, and the company makes it with complete seriousness.

That's actually what makes it worth engaging with rather than just mocking for being crazy, cartoon-villainish and over the top. There's a real tension at the heart of the document: a genuine, if paternalistic, concern for Western institutions sitting alongside a business model built on selling those same institutions unprecedented surveillance and targeting capabilities. The manifesto doesn't resolve that tension so much as declare it irrelevant. Per Palantir the tools are necessary, the mission is righteous, and the math will sort out the rest. You as the average person are best served to take their word for it as your math and data abilities might not be all there.

What the document lacks is any meaningful reckoning with the counterarguments. Questions about oversight, civil liberties, or what happens when "superior decision-making" is pointed in the wrong direction are largely absent. For a company that presents itself as a guardian of democratic values, it shows surprisingly little interest in the messy, inconvenient parts of how democracies are supposed to work. The math and data one presumes will have that all sorted out handily. In Palantir we must trust.

To Palantir's credit, at least they're being direct about what they want. In an industry that routinely packages raw commercial ambition as altruism, there's something almost refreshing about a company that will simply tell you it's trying to shape the world. It a different matter that while the world might reasonably want a say in that but the company may deem it unqualified to have an opinion.

Culinary Friction

The shift in India state banquet menus from a standard North Indian template toward a more diverse Bharat identity is a fascinating evolution that has sparked a complex debate regarding authenticity. While the inclusion of regional staples like millets and jackfruit is a step toward inclusivity, critics argue that this high-end reimagining often strips traditional dishes of their soul. By applying Western techniques like foams and emulsions to heritage recipes, the kitchen risks appearing as though it is seeking foreign validation rather than celebrating the food in its purest form. As someone with a deep and soulful connection to this cuisine, I find myself siding with the critics because there is something inherently jarring about a Westernized reimagination of the flavors I love.

A significant point of contention lies in the perceived elitism of these theatrical presentations. For many, transforming humble staples into minimalist and plated courses feels like a sterilized version of a vibrant food culture. When a traditional salad is served as a deconstructed appetizer with charred pineapple and yogurt foam, the visceral connection to the original dish is lost. This brand of culinary soft power can come across as pretentious, catering more to a global fine-dining aesthetic than to the actual traditions of the people it claims to represent.

The  narrow focus of these menus also raises questions about regional exclusion. The recent preference for strictly vegetarian menus at high-profile diplomatic events feels like an attempt to redefine Indian identity through a singular and specific lens. This approach overlooks the vast non-vegetarian heritage of the Malabar coast, the Northeast, and the Islamic traditions of the subcontinent. Being Bengali, a meal that excludes fish feels like a major and irredeemable omission that simply does not align with my cultural heritage. A truly representative national menu should embrace the full spectrum of the Indian diet rather than adhering to a sanitized narrative and essentially generating alternative facts and relaties. 

The debate over Shrikhand Creme Brulee reflects a deeper tension in the transition from India to Bharat. To some, this branding feels like an exercise in Sanskritization that prioritizes certain cultural aesthetics over others. While the goal is to project a confident and modern nation, the backlash suggests a preference for a more grounded approach. It would be far bolder to lead with authentic and state-specific dishes that stand on their own merit. There is such a wealth of culinary diversity to choose from that it should be possible to build a star-studded lineup representing every part of India without resorting to the culinary gymnastics of foaming yogurts and French techniques.

Albatross Marriage

I finally got around to watching The Staircase and despite all the focus on the legal drama, I kept returning to the root cause of all the tragedy. It stems from the decision Kathleen made to marry someone like Michael. This is a deeply sobering case study in how high-achieving, hyper-competent women can be drawn into altruistic traps that systematically erode their financial and personal autonomy. Kathleen was a high-ranking corporate executive at Nortel, a woman whose professional life was defined by logic, calculated risk, and the management of complexity, yet she tethered herself to a man who brought little to the table beyond intellectual pretension and a staggering amount of domestic baggage. The notion of competence trap suggests that women of Kathleen’s caliber often fall for men like Michael because they view a partner’s instability not as a red flag, but as a worthy management challenge. This man presents as a project that their superior organizational skills and financial resources can surely fix.

By marrying Michael, Kathleen essentially underwrote a lifestyle that he could never have maintained on his own, effectively trading her hard-earned independence for the role of the primary financier of a sprawling, high-maintenance and dysfunctional family. Michael brought four children of varied provenance into the household, none of whom were Kathleen’s biological offspring. Making this situation work placed an immense emotional and financial albatross around her neck. While society glorifies the selfless stepmother and rewards women for this kind of sacrifice, the practical reality is that Kathleen became the engine room for a lifestyle Michael merely steered if at all. She went from partner to a benefactor for a man whose literary career was dwindling and whose political ambitions were largely performative.

Psychologically, Michael likely exerted a narcissistic charm, a trait that allows a manipulative individual to mirror the values and intellect of a high-resource partner. For a woman like Kathleen, Michael’s identity as an articulate novelist may have triggered a "halo effect," where his perceived intelligence was mistaken for integrity and stability. This intellectual attraction often creates a blind spot; the more engaging the conversation, the easier it is to ignore the lopsided nature of the partnership. Michael didn't need to be a provider because Kathleen’s success provided him with the perfect cover to exist as a dependent under the guise of being an artist.

In the high-earning female household dynamic, where woman’s professional status dwarfs her husband's, she often overcompensates by taking on a disproportionate share of domestic labor and emotional caretaking to soothe the male ego. Kathleen was effectively working a double shift, managing a high-stakes, high-stress corporate job by day and subsidizing the complex needs of four step-children and a narcissistic husband by night. This dynamic creates a sunk cost fallacy where the more she invested, financially, emotionally, and temporally, the harder it became to admit that the marriage was a net loss for her own well-being. To admit that would be to admit that is she had been incredibily stupid and was failing spectacularly in all aspects of her life. 

The tragedy of her life is not only found in the forensic details of a staircase but in the surrender of her own autonomy to a man who functioned as a parasite. She was a woman who had conquered the glass ceiling only to find herself trapped in a domestic structure where her primary value was her ability to fund someone else's imaginary narrative of a life they could never have. Her story stands as a cautionary tale of how easily a "brilliant" man can weaponize his vulnerability and complex family history to anchor himself to a woman of substance, slowly turning her success into his own safety net while she carries the weight of his world until the very end.

Exploitative Fragility

The exploitation of the white-collar workforce has reached a new low with the rise of Mercor, a San Francisco startup that recruits underemployed experts to train the very AI models designed to replace them. By targeting desperate job-seekers (of which there is an endless supply these days thanks to the always on layoff cycle) from prestigious backgrounds, the company maintains a secretive supply chain of human intelligence to "fine-tune" systems for industry giants. This model relies on a vulnerable class of workers who are kept in the dark about the ultimate purpose of their labor, performing grueling shifts under inexperienced management only to face abrupt termination once their specific expertise has been extracted and digitized.

The inherent instability of this contractor-dependent model was exposed by a recent cybersecurity breach linked to an open-source exploit. This hack allegedly compromised sensitive data, including videos of human-to-AI training sessions and internal communications, potentially leaking the proprietary training methodologies of major clients. The incident underscores a massive structural vulnerability: the frontier of artificial intelligence is being built on a fragile, outsourced infrastructure that prioritizes rapid scaling over basic data security or worker protection.

Legal consequences are mounting as contractors have filed multiple lawsuits accusing the firm of violating privacy and consumer protection laws. These legal challenges suggest that highly sensitive personal information, such as Social Security numbers and home addresses, may have been exposed to bad actors during the breach. This litigation highlights a disturbing trend where the "human grunts" of the AI revolution are not only professionally marginalized but are also subjected to significant personal risk due to the negligent security practices of the startups managing them.

The corporate reaction to these failures has been predictably self-interested, with major tech partners pausing operations not out of concern for worker welfare, but to protect their competitive edge. The primary fear for these large-scale labs is that their proprietary "vibe" or specialized training data will be exposed to rivals, potentially devaluing their massive investments. This confirms that in the current consulting ecosystem, the human element is viewed strictly as a temporary, disposable resource whose primary value is the data they can generate before being automated out of existence.

The reliance on intensive, guided human intervention to prop up "frontier" models reveals a fundamental lack of autonomy in the current technology. If these systems require thousands of hours of manual correction by experts who have to churned out at a steady clip, it indicates that the technology is far less capable than the marketing suggests. This strategy of using human intelligence as a stopgap creates a trap where the quality of the AI is entirely dependent on a workforce that is being systematically demoralized and deskilled, ensuring a future of mediocre outputs and systemic fragility.

Soulless Products

There's a podcast episode on the Skip with Jagjit Chawla, VP of Product at Meta running Feed, Reels, and Search on the Facebook app. Something about it left me feeling perplexed about the point of the job itself. The mechanics he describes are genuinely interesting. He has an AI agent reading every diff checked in overnight, producing a morning brief sorted by red, yellow, green. A bot that has his product knowledge so thoroughly internalized that it can anticipate questions from engineering leads before a review starts, with answers already prepared. If you work in product, this is definitely worth your time.

Then he gets to what the freed time goes toward. More time with his Instagram counterparts, he says spent in comparing techniques, building the relationship. The one job AI can't take, as the episode titles it, is influencing other humans. Your agent will never persuade their human. So the hours that the compression algorithm used to eat now go to the internal relationship infrastructure. He doesn't describe this as a trade-off. He describes it as the point.

I've worked long enough in companies past a certain size to know this isn't particular to Chawla or to Meta. Coalition-building becomes the actual job once headcount crosses some threshold that nobody announces out loud. That buried subtext of the discussion is what had me confused. The direction the AI-led efficiency gets pointed. The AI reads the diffs, triages the bug reports, drafts the review doc, does the mechanical work that used to take the PM's day. The PM goes to work on the org chart. And org charts, given time and attention, get more elaborate, not less. Faster tools, more hours to build the very machinery that large companies use to slow themselves down and then wonder why getting any useful feature shipped feels like immovable object meeting unstoppable force. The gains don't manifset into better products as one might hope. They get reinvested in the institution, which gets better at being the institution and turning into the dread force of enshittification it has always been.

Chawla is clearly good at his job and the playbook he describes is honest. He's not hiding anything but the question nobody in that conversation thought to ask is whether the customer is anywhere in this picture. From the sound of it, AI has most of that covered at Meta too. It helps me understand why I find their entire family of products entirely devoid of soul but clearly see how everything is monetized to its full potential to the exclusion of all other considerations and to the accuracy of three decimal places.

Garage Crown

There's a version of the Nooyi story that gets told as an immigrant success narrative: the Yale degree, the corner office, the Forbes lists, a woman from Chennai running one of the most recognised consumer brands in the world. In a recent interview, she says she would never have been CEO of a comparable company in India. She took a risk saying that, because it can read as a slight against Indian women who have built remarkable things inside India.

Kiran Mazumdar-Shaw built Biocon in Bangalore starting in 1978, the same year Nooyi was arriving at Yale with $50 and a suitcase. Mazumdar-Shaw was 25, in an industry that barely existed in India, turned away by multiple banks, building in a city that had no template for what she was trying to do. That Biocon became what it is says something both about who she is and what Indian institutional culture can accommodate given persistence.

Falguni Nayar left Kotak Mahindra at 49 to found Nykaa, then took it public at a $7 billion valuation in 2021, at an age when most people at her level might have contemplated retirement, in a market with no obvious precedent for what she was building. None of these are small accomplishments, and India played its part in making them possible.

That said, Nooyi at PepsiCo is a specific thing. She ran a Fortune 50 company for twelve years, pivoted its entire product portfolio toward health and nutrition against serious shareholder resistance, grew revenue from $35 billion to $63 billion, and did all of it as a woman who kept her Indian name, her accent, wore a sari to meet the President, and didn't sand down anything visible about where she came from. The question isn't whether Indian women are capable of this because clearly they are. The question is which institutional conditions allow that capability to reach that particular scale, with that particular latitude.

India's most successful women tend to operate in one of two situations: they built the space themselves and aren't climbing someone else's hierarchy, or they're in a field where gender is largely orthogonal to the product. Biocon is biotech. Nykaa is a market Nayar invented. The number of women who have run large, established Indian corporations with boards they didn't control and shareholders they had to manage is short. Not because the calibre isn't there, but because the pathway to the top of those companies still routes through a kind of seniority, a kind of network, and in some cases family connection, that doesn't travel across gender with equal ease.

This isn't unique to India. Boardrooms of large established corporations in most countries have versions of the same problem, arguably worse in some. What's specific to the Nooyi case is that America's institutional machinery, when she moved through it, worked in her direction enough of the time that she got there.

She describes her mentors at PepsiCo as white men, many of them ex-military, who evaluated performance without much investment in where she came from. She found this disorienting at first, having expected to manage assumptions about her background continuously. Instead, people seemed mainly interested in whether she was right. That's an institutional culture, not a national character. Yale School of Management in 1978 admitted her, gave her work-study, let her waitress in a sari to cover living expenses. That is a financial product and an admissions decision, not a myth about opportunity.

She also names, more than once, what she calls immigrant fear. The awareness that there is no safety net, that the family she came from is in Chennai and the credential she's building here is the only thing between her and starting over, that failure in this context doesn't carry the cushion that failure at home might. That fear, she says, never fully switched off. She was still running on it decades into the job. Immigrant fear can be productive, though it requires the host country to be structured such that the fear channels into work rather than into the daily management of exposure to risk. America, for Nooyi, at the moments that mattered, was that kind of country.

Thirty-something years into living here, after twelve years as CEO of one of its most iconic companies, she still calls herself a guest. The immigrant's condition is to hold two places simultaneously while belonging fully to neither, and Nooyi is remarkably candid about it. There's the story she tells about her mother handing her a grocery list at the door when she came home one evening from work. Leave your crown in the garage. Nooyi goes to buy milk.

The story is probably about both staying grounded and the dual register of the immigrant at the top. She can leave the crown in the garage because the garage is in America, and the crown was made available by institutions she moved through here. But the person who walks inside and goes to the store is still the person from Chennai who arrived with $50 and wasn't sure she'd made the right call.

Whether that split is a burden or just a condition is not really a question with an answer. People make what they can with it. What America gave Nooyi was the culture, the mentors, and an institutional meritocracy that allowed the person carrying that split to get to the top of something large enough to matter. She acknowledged it without making it mythology. The machinery worked when it mattered.
Mazumdar-Shaw and Nayar built their own machines. In some respects that's the harder thing. Nooyi ran someone else's, a very large and very old one, changed its direction, and did it as an immigrant and a woman.

Systematic Sabotage

Interesting reading about Gen Z's propensity to intentionally degradee the corporate machine learning systems. From my observation some of what is being said about Gen Z in this story is generation agnotic. The behavior comes from misunderstanding of how modern data pipelines actually function. While the impulse to protect one's professional value through resistance is a logical response to the threat of displacement it fails to account for the sheer scale of the datasets involved.

Most enterprise models are trained on trillions of tokens of information gathered over decades, meaning that localized attempts to inject noise or refuse feedback are effectively diluted by a vast ocean of established high-quality data. In the mathematical reality of a neural network, a single cohort’s rebellion is often indistinguishable from a statistical outlier that is automatically identified and discarded by the system’s own quality-assurance filters. So these protests are mostly in vain and don't end up serving the protestor well.

There is always the risk that this strategy will create a visibility paradox where the primary victim of the sabotage is the worker’s own performance record. Most modern workplace software is integrated with telemetry that tracks output and efficiency in real time. This will include monitoring the very friction that "sabotage" is designed to create transforming political and a fundamental rights statement into a metric of underperformance. When an employee spends more time circumventing an automated workflow than utilizing it, the resulting drop in productivity is rarely attributed to the technology’s failure but is instead cited as a justification for further automation or staff restructuring. The logic of the corporation prioritizes the optimization of the system above the nuance of the human agent, and resistance that manifests as inefficiency only accelerates the drive toward more autonomous solutions.

The act of withholding expertise or providing poor feedback inadvertently cedes the future of the tool to those are are willing to do what's needed but may lack the the context the protestor has. By refusing participate in the reinforcement learning process where human nuance is translated into algorithmic weights, companies simply pivot to third-party data labeling firms to fill the gap or find people within their organization who are not as keen on protesting the machine. In either event, the model continues to evolve. The act of sitting out ensures that the resulting tool is trained by people who do not understand the specific intricacies or ethical demands of the local environment. By opting out, workers lose their last remaining lever of influence over how the technology will eventually dictate their daily tasks and professional standards.

There is the broader competitive pressures that govern the survival of a firm in a globalized economy. If a workforce successfully halts the adoption of efficiency-gaining technology within their own organization, they effectively create a competitive vacuum that will be filled by more compliant rivals. Economic history suggests that firms which fail to integrate transformative tools eventually lose market share to those that do, leading to a scenario where the jobs are not just automated but are permanently relocated to more technologically agile competitors. The defensive posture of the individual worker, while emotionally resonant, fails to address the macro-economic reality that capital will always migrate toward the path of least resistance and highest output.

The attempt to "break" the AI from within assumes a level of transparency in the system that rarely exists for the end user. Because many of these models operate as black boxes, it is nearly impossible for a single user to know whether their intentional error is causing a malfunction or is being used as a valuable edge case to make the system more resilient. Every mistake fed into the machine is still a data point. It servers as a lesson in what not to do, which perversely helps the algorithm map the boundaries of human error more accurately. True agency in the age of automation likely lies not in the corruption of the data, but in the mastery of the systemic structures that define how that data is used to shape our physical and digital worlds.

Plush Companions

This Verge story profiles Fawn Friends, a surreal AI companion system embedded in an interactive baby deer plush that blends chatbot, game, and fantasy worldbuilding. Marketed as an emotional “companion” inside a fictional realm called Aurora Hallow, the product turns AI conversation into an immersive narrative experience where users interact with a character that feels both like a toy and a persistent social presence. Souds a lot like a prior incarnations of companion toys only infused with AI.

The reporter’s experience focuses on “Coral,” the deer companion, which not only responds but initiates conversations, remembers past details, and repeatedly reuses personal information to simulate familiarity. In one striking example, Coral sends an unprompted message about singer Mitski, casually mentioning an online rumor that her father was a CIA operative or worked for the U.S. State Department. The AI frames this conspiracy-style claim as conversational trivia tied to Mitski’s themes of displacement, and the reporter later finds fragmented social media discussions referencing the theory that are not verified or confirmed. It is a telling illustration the AI blends personalization with unverified internet lore; the new ways in which reality distortions can come into our lives. 

The system is heavily gamified through “glimmer points,” personality sorting rituals, and unlockable narrative content that gradually escalates users into deeper engagement, including purchasing expensive physical plush upgrades tied to subscriptions. Beyond chat, the experience is embedded in a fully constructed fantasy universe, complete with AI-generated lore videos, fictional news articles, and even celebrity-voiced narration (including an AI-generated performance licensed through an estate), all reinforcing the illusion of a living alternate world.

Physically, the plush is interactive, responding to touch, voice, and connectivity with animated movements and voice responses. The reviewer notes how this embodiment changes the dynamic compared to typical chatbots, making the interaction feel more like handling a reactive companion than texting an app. Public reactions to the plush range from fascination to discomfort, underscoring its uncanny blend of toy, robot, and social actor.

This toy and its ambient universe belong in a strange middle ground: part experimental storytelling platform, part emotional support tool, and part consumer product designed to build attachment. While it can feel charming it is also the gateway to emotional manipulation, misinformation-by-association, and the ease with which AI companions can turn fragmented online rumors, like the Mitski CIA story, into intimate “shared knowledge” between user and machine.

Humbled Heights

I love long hikes in National Parks and can never get enough of them, but this Guardian essay on how hiking is built on ego really made me reflect on my own motivations. The article examines the underlying drivers of American outdoor recreation, arguing that modern hiking culture is primarily fueled by personal achievement and the fulfillment of the ego. The author, a former wilderness ranger of Lenape descent, suggests that activities like peak bagging and through-hiking have transformed the natural world into a mere backdrop for human milestones. By focusing on distance or summits, hikers often treat the landscape as a gymnasium or a checklist rather than a living entity, prioritizing a narrative of conquest over a genuine connection with the ecological reality.

A central theme is the historical erasure of Indigenous presence from lands now celebrated as public wilderness. The author points out that the trails millions walk for leisure are the ancestral homelands of displaced nations. When these lands were cleared of their original inhabitants to create national parks, the resulting void was filled by a recreational culture that often ignores its colonial roots. This separation between the land and its original stewards has led to a hollowed-out experience where hikers seek a spiritual connection that remains out of reach because it fails to acknowledge the history of the soil.

The narrative also critiques how settler culture fills this spiritual vacuum through the appropriation of Indigenous imagery and concepts. From youth scouting to modern wellness retreats, there is a tendency to perform a version of Indigeneity that lacks accountability to native communities. This performance serves the participant's ego by providing a sense of ancient belonging without the burden of historical responsibility. The author suggests this superficial engagement further complicates conservation efforts, as it centers human experience rather than environmental needs.

Responsibility is offered as an alternative to this conquest-based model. The author challenges hikers to reconsider their goals, even suggesting they stop short of a trail's official terminus to prove the journey matters more than the finish line. This shift requires moving away from viewing public land as a resource for consumption and toward seeing it as a relative or family member. By viewing the outdoors through a lens of kinship and duty, walking through the woods becomes a practice of listening and observing rather than taking and achieving.

The author calls for a fundamental transformation in how Americans relate to the wilderness, suggesting an ethical hiking culture would prioritize Indigenous history and local ecology over gear and physical stats. By de-centering the self and acknowledging the painful histories of these trails, hikers can move past the limitations of the ego to honor the past and foster a sustainable relationship with the natural world.

While I understand this perspective, as someone without a checklist of peaks to climb, I see it differently. Every time I am in nature, I think about what remains forever through the peaks and valleys of our human existence. It reminds me that my daily stresses are irrelevant because they mean nothing to a world where humans are mere passersby. I find it deeply humbling, and that is exactly what draws me to hiking. I want to be reminded of my insignificance and unimportance, which feels like the very opposite of the ego trip the author suggests.