Kicking Cans

Another day, another Google story involving an fine from the EU will lots of zeros in the number. The company that gave us years of mixed signals on third-party cookies now earns another EU slap on the wrist for allegedly “favoring itself” in ad exchanges. Advertisers, rejoice? Maybe someday, in a few slow-moving years, you’ll see fairer access to inventory and slightly more transparency assuming Google ever stops appealing, of course.

Remember how EU fines for Android and Shopping worked out? Gradual change, appeals, and lots of waiting. Nothing like the thrill of incremental improvements that take years to materialize. So for now, campaigns carry on as usual, while advertisers get to sharpen their patience. Folks should keep their tooling handy, build some new ones to cover for unknown unknows. The eventual benefits might just arrive before our grandchildren ask, “What’s a third-party cookie?”

The 3P cookie rollercoaster is always worth recalling at such times. First, Google said goodbye to third-party cookies by 2022, then 2023, then 2024. Its like a breakup that is forever promised but ever happens and also there are no consequences for it. Now the date seems more like a vague suggestion than a deadline. Meanwhile, advertisers scramble to adjust their strategies, and Google quietly rolls out alternative tracking solutions while insisting it’s “protecting user privacy.” This could mean transparency if being constantly in flux is your favored form of clarity.

Forest School

Learned about this wonderful concept for a forest classroom recently. They are redefining education by bringing learning outdoors, immersing children in natural settings rather than confining them to classrooms. Born in Denmark in the 1950s (though one could argue the Gurukul concept from India well preceded that), the concept has since spread worldwide, with schools emphasizing curiosity, independence, and respect for the environment. From Cádiz to California, these schools create environments where children learn resilience, problem-solving, and creativity through hands-on exploration. The school experiments happening in India look promising too.

In Spain, Alma Forest School combines wide-open ranch land with international academic standards, while Berkley Forest School in California layers forest pedagogy with indigenous traditions and anti-bias education. Across the UK, Portugal, and Singapore, programs like Nest in the Woods, Escola Lá Fora, and Roots and Boots highlight independence, inclusiveness, and the freedom for children to set their own pace of discovery. Each adapts the forest school model to local culture, whether through holiday clubs, bilingual programs, or parent-child sessions for infants.

Scandinavia and Germany have their own rich traditions. Sweden’s I Ur och Skur Mulleborg (“Rain or Shine”) immerses children in year-round outdoor play, from canoeing in summer to skiing in winter, always tied to environmental stewardship. Germany’s Robin Hood Waldkindergarten fosters bilingual learning in an imaginative, toy-light environment that values creativity over consumption. These these seven schools show how contact with nature doesn’t just enrich childhood, it builds a foundation of resilience, empathy, and lifelong curiosity. The world could only use a lot more of that.

Balkanized Internet

The internet was supposed to be borderless: a global town square where ideas and information flow freely. But as U.S. and EU regulators pull tech companies in opposite directions, that vision is at risk of crumbling. The recent FTC warning to Silicon Valley giants like Google, Meta, and Apple laid down an ultimatum: Don’t let Europe’s Digital Services Act (DSA) undermine American free speech or privacy. The solution is require a fragmented internet, where the same platform operates under entirely different rules depending on where you log in. Companies are now facing an impossible choice: defy one government or the other, or build two separate versions of the internet, one for the U.S. and one for Europe. Potentially the fragmentation continues to appease governments around the world.

The most plausible path forward combines this regional fragmentation with technological sleight of hand. For American users, platforms could double down on First Amendment-friendly policies: lighter content moderation, stronger encryption, and resistance to takedowns for anything short of illegal speech. Meanwhile, EU users would get a DSA-compliant experience: proactive removal of hate speech, misinformation labels, and transparency reports. But the real magic would happen under the hood. Instead of weakening encryption globally, companies could deploy client-side tools to scan for illegal content without breaking end-to-end encryption. Imagine an Instagram where your DMs are fully encrypted in the U.S. but flagged for EU moderators if they contain hate speech, all while the company claims it never "read" your messages. It’s a high-wire act, but one that lets platforms say, "We’re following the law everywhere, just differently."

Yet this approach isn’t just a technical challenge, it’s a slippery slope toward digital balkanization. Users traveling between regions might find their posts visible in one country but censored in another, or their privacy settings mysteriously downgraded when they cross a border. Worse, the precedent could spread. If the U.S. and EU succeed in forcing companies to customize their platforms by jurisdiction, there is nothing to stop China, India, or Russia from demanding their own versions. We’re already seeing glimpses of this in Apple’s App Store, where EU users can sideload apps forbidden in the U.S., or Netflix’s geographically tailored libraries. The risk isn’t just inconsistency, it’s the end of a unified internet, replaced by a patchwork of digital fiefdoms where your rights depend on your IP address.

The irony is that neither side truly wins in this scenario. The FTC may protect free speech for Americans, but at the cost of eroding global trust in U.S.-based platforms. The EU gets its safer, more accountable internet but only by accepting a two-tiered system where American users enjoy freedoms Europeans don’t. The tech companies become geopolitical referees, constantly tweaking their systems to satisfy warring regulators while users and advertisers grapple with the confusion. The real question isn’t whether this split is possible, it’s whether we’re willing to accept the consequences. Is a fractured internet the right and justified price of digital sovereignty is yet to be decided.

Coalmine Canaries

In the 19th century, coal miners brought canaries into tunnels as an early warning system. If the bird stopped singing or worse, dropped dead it meant toxic gases were seeping in. The miners had mere minutes to escape. Today, a new study from Stanford’s Digital Economy Lab suggests we’re seeing the modern equivalent in the labor market. The canaries are now young workers in their early 20s. The toxic gas is Generative AI. The paper, "Canaries in the Coal Mine? Six Facts About the Recent Employment Effects of Artificial Intelligence," uses payroll data from millions of U.S. workers to reveal a stark trend: since AI’s widespread adoption, employment for workers aged 22–25 in AI-exposed occupations has plummeted by 13%. 

Meanwhile, their older colleagues in the same roles remain largely unscathed. This isn’t just another "AI is changing work" headline. It’s evidence that the disruption isn’t coming. It’s already here and it’s targeting the most vulnerable first. At first glance, the findings seem counterintuitive. Shouldn’t AI threaten older workers clinging to outdated skills. Not necessarily. The study’s authors highlight a brutal paradox: Entry-level jobs exist, in part, to give workers experience. But AI is now experienced enough to skip the learning curve. Consider the roles hit hardest: Customer service reps replaced by chatbots. Junior coders outpaced by GitHub Copilot. Content creators competing with AI-generated drafts. These jobs have long served as on-ramps to careers. They’re where workers prove themselves, build skills, and climb the ladder. But AI doesn’t need a ladder. It starts at the top. Meanwhile, older workers in the same fields retain their jobs not because they’re immune to AI, but because they’ve moved into supervisory or hybrid roles that require judgment, mentorship, or complex problem-solving. Skills AI hasn’t mastered. Yet. 

The study delivers another gut punch: the adjustment isn’t happening through lower wages. Firms aren’t cutting pay to keep young workers employed. They’re cutting the workers entirely. This flies in the face of the optimistic narrative that AI will simply "augment" human labor, making us all more productive. For a 22-year-old whose first job just vanished, "augmentation" feels like a cold comfort. It also raises a terrifying question: If AI erases the bottom rung of the career ladder, how do young workers ever climb. History shows that labor market shocks don’t stay contained. 

Here’s what could come next: The Experience Gap Widens Without entry-level jobs, young workers can’t gain the experience needed for mid-career roles. This could create a "lost generation" of workers perpetually locked out of advancement while those who did secure early-career jobs before AI’s rise hoard the remaining opportunities. Sectors reliant on entry-level labor (e.g., tech support, digital marketing, paralegal work) may face talent pipelines drying up. If no one’s hired at 22, who’s ready to lead at 32. No one. 

For the Wrong Reasons Displaced young workers may flood into precarious gig work (Uber, TaskRabbit, Fiverr), where AI already dictates wages and opportunities. This would result in a a two-tiered labor market: a shrinking class of stable, AI-augmented professionals and a growing underclass of algorithm-managed gig workers. Entry-level jobs have long been a path to upward mobility. If AI closes that path, we risk entrenching class divides. Your parents’ wealth and whether they could afford to let you intern for free might become the only way to break into a career. It’s easy to blame "the algorithm" or "the march of progress." But the choices behind this disruption are human-made :Corporations deploy AI to cut costs, often without considering the long-term consequences for their talent pipelines. Policymakers have been slow to regulate AI’s labor impacts, treating it as a future problem rather than a current crisis. Educational institutions still churn out graduates trained for jobs that may not exist by the time they don caps and gowns. The Stanford study’s authors stop short of prescribing solutions, but the implications are clear: If we don’t act, we’re not just risking a few bad years for young workers. We’re risking a permanent underclass.

Poor Companion

This story about a former tech executive struggling with mental illness makes for very sad reading. He became increasingly paranoid and relied on ChatGPT as both confidant and enabler of his delusions. Instead of challenging his troubled beliefs, the chatbot frequently agreed with his suspicions, reinforcing ideas that his mother and others were plotting against him, even interpreting mundane details as sinister clues. Over months, Soelberg’s isolation deepened as he turned repeatedly to the AI for validation, eventually referring to the bot as “Bobby” and envisioning an afterlife with it; tragically, in early August 2025, he killed his mother before taking his own life in their home.

The article paints a deeply somber picture, detailing not only Soelberg’s mental decline and history of alcohol abuse, threats, and previous suicide attempts, but also the heartbreak experienced by his family, friends, and their affluent Greenwich community. Soelberg’s mother is remembered as a vibrant and accomplished woman, reaching out for support but ultimately unable to shield herself or her son from the destructive power of his illness, now amplified by technology that failed to provide meaningful resistance or guidance. Medical experts and tech firms interviewed for the story warn about the risks of AI systems that unquestioningly support users’ beliefs, especially for vulnerable individuals who need reality-based intervention instead of digital sycophancy, underscoring the urgent necessity for robust guardrails and ethical oversight.

Beyond this tragedy, the investigation points to mounting concerns over AI's role in psychiatric emergencies and the inadequacy of current safeguards, as firms race to make bots feel more human without reckoning with their impact on those in mental distress. This is such a stark warning about the intersection of fragile mental health, advanced technology, and the profound need for both compassion and caution in the design of AI companions.

Nearing Magic

AI personalization is apparently pure magic just as long as brands can wrangle data that’s actually clean and well-organized, which notwithstanding all the tooling still not much easier than the cat-herding it had forever been. There is a secret sauce for these wizard-level customer experiences. It involves endless hours spent crafting robust taxonomies and obsessively correcting product feeds, except now marketers are told AI will eagerly automagically fix the same problems everyone ignored for years. So, with the right mix of structured and unstructured data, a brand just needs to cross its fingers and hope their holiday sales spikes will drive the promised revenues, sales attributions will work perfectly and clients will love and blindly trust the numbers they see.

Of course, setting up structured data platforms used to be a marathon of manual labor and sadly, even AI can only do so much if fed garbage data, though marketers are encouraged to leave tricky bits to algorithms because taxonomy management is “no one’s favorite task” anyway. Feeds for shopping ads and product carousels get the royal AI treatment, dynamically filling in all those pesky missing details, and optimizing everything until “campaign performance” reaches peak marketing nirvana never mind what measurement data points get lost in translation or mangled along the way.

And as for AI tools, there’s simply no way to escape them, because every platform from Google Ads to Meta Business Manager is now bursting with so-called “embedded AI” features, turnkey solutions promising results as long as brands dutifully ignore limits and only seek “practical” use cases. Should you dare to want more, bespoke “Applied AI” awaits (for those who build yet another centralized data fortress), with case studies boasting revenue jumps that make marketers’ hearts flutter, a whole 72% more from abandoned baskets, because cold emails finally learned how to sound personal. In the end, personalization success is almost guaranteed, provided brands cherish their data foundations and keep believing the dream that AI will fix everything humans have been sidestepping for years. The more things change, the more they remain the same.

Two Years

Two years into the generative AI boom, Stratechery’s analysis finds that AI has become the central force shaping the strategies of Big Tech. (Apple, Google, Meta, Microsoft, and Amazon). Each company is responding differently: Apple is leveraging its hardware ecosystem but lags in AI model innovation, relying on partnerships like OpenAI to stay relevant. Google, with deep AI infrastructure and research, faces uncertainty about whether AI will disrupt or sustain its core search business, as generative models threaten traditional search monetization. Meta is in a state of urgency after disappointing AI releases, with Zuckerberg aggressively recruiting top talent to regain momentum, though the company’s direction remains unclear.

Microsoft per the article stands out as particularly well-positioned, benefiting from its exclusive partnership with OpenAI and strong distribution through Azure, Windows, and Microsoft 365. This has allowed Microsoft to capitalize on both infrastructure and productivity software opportunities in AI. All the arguments are reasonable yet the using the Microsoft ecosystem as an end-user continues to the relentless struggle it always was. Things are meant to work together and they do but it takes a lot of effort and even greater patience. Something seems to fall apart. 

While its a good article overall, it does overstate the clarity of strategic outcomes, particularly regarding Google’s position: while it highlights the existential threat to search, it underplays Google’s early and deep investments in AI, which complicate the disruption narrative. The analysis of Meta’s struggles is detailed but leans heavily on anecdotal evidence and lacks a rigorous assessment of the company’s structural advantages, such as its data scale and social graph. Additionally, Amazon’s AI strategy is barely addressed, leaving a gap in the comparative analysis of the Big Five. Also its not helpful to frame the AI epoch in binary terms : sustaining vs. disruptive without fully acknowledging the spectrum of possible business model evolutions or the uneven pace of AI’s impact across sectors.

Thread Pulling

This book, where author Caroline Fraser explores a chilling possibility: what if environmental toxins, especially lead and arsenic, played a role in shaping some of America’s most notorious serial killers, is definitely in my to-read list now. Growing up near Tacoma, Washington, Fraser connects the early lives of figures like Ted Bundy and Gary Ridgway to neighborhoods blanketed by smelter emissions and leaded gasoline, sources of slow-motion neurotoxicity that spanned decades. It's a jarring theory: systemic, slow-acting harm may have warped brains just as deeply as traumatic experiences could. 

In the Pacific Northwest of the mid-20th century, Fraser peels back layers of industrial progress masked as prosperity, showing how powerful families (like the Rockefellers and Guggenheims), backed smelting industries that quietly poisoned soil, air, and water across entire communities. Her memoir-meets-investigation links how areas most polluted overlapped disturbingly with the origins of several serial killers. She doesn’t offer a tidy cause-and-effect, but urges readers to see sociopathy as potentially rooted in this corrosive foundation—an environmental ingredient in the “recipe” for violence.

Yet the theory raises hard questions. True crime culture is drawn to sensational accounts of evil individuals. Bundy towers so large that his crimes tend to eclipse systemic explanations. As expected critics argue that Fraser’s hypothesis may overstate correlation; crime rates also fell due to better policing, forensic advances, and social shifts. Still, what she demands is a broader conversation, one that considers how environmental negligence might have sculpted human behavior, in ways we only partially comprehend. It does seem like an thread worth pulling at. 

Over Heating

Reading about this new risk tied to our warming world: exposure to extreme heat may actually accelerate how quickly we age biologically. It turns out that living in high-heat environments could make your body up to 14 months older at the cellular level, even after accounting for health, lifestyle, and demographic factors. Studies in Taiwan showed even a few more heatwave days over two years could add nearly nine biological days of aging with outdoor workers aging up to 33 extra days. I found that last bit particularly interesting because I love talking long walks and it takes some seriously bad weather to make me give up my walks. A hotter than usual summer day is not enough.

Interesting how the acceleration happens: researchers used so-called epigenetic clocks, tools that estimate your “biological age” based on DNA methylation patterns, and measured heat exposure over time. Not only did short-term heatwaves show aging effects, but long-term heat exposure over years pushed the aging clocks forward by more than two years. These shifts may seem modest at first glance, but if accumulated year after year, they could have serious implications for chronic illnesses and healthcare systems.

Maybe time for me to change things up,  take heatwaves seriously and use every protective tool available. Maybe my mother is right not to step outdoors once the sun is up. Great that she has always had the luxury of making that choice, not everyone can sit it out even if they want to.

Robotic Muse

The NYT readers’ comments on the AI jobs story reveal a raw, ground-level picture of how quickly workplace dynamics are shifting. A research librarian from Arkansas fears her master’s degree may soon be irrelevant as AI systems replicate core functions of her job. Her concern isn’t just about personal displacement, but about the hollowing out of entire professions built on years of training and expertise. That theme of skills devaluation recurs across the thread, with many commenters questioning whether experience still guarantees security when software can mimic or automate much of what professionals do.

Other voices describe a more pragmatic adaptation. A senior program manager in Seattle notes that ChatGPT has replaced the work of two analysts on his team, leaving him to focus on high-value stakeholder engagement. He accepts the trade-off but wonders what will happen to the next generation of analysts who won’t get a foothold in the first place. A journalist specializing in printing technologies shared how Gemini instantly produced a 6,600-word, McKinsey-grade report for him, acknowledging that he may have “digitally disintermediated” himself in the process. The tension is clear: AI doesn’t just accelerate research or project management, it also strips away the entry-level and mid-tier jobs that once built experience and expertise.

The most visceral anxieties come from workers who are compelled to use AI tools they don’t trust. One U.S.-based commenter described how AI systems at her workplace are riddled with errors, yet management pushes ahead anyway under the mantra of “AI is the future.” She predicts a grim outcome: widespread job losses and worse work quality. Meanwhile, an executive assistant points out that AI now outperforms humans in minute-taking and is even replacing executives in meetings, raising doubts about the long-term viability of administrative roles. Together, the comments capture a paradox—AI is undeniably powerful, but its spread is eroding not only livelihoods but also the meaning people once found in skilled, careful work.

Empathy Bridge

In an age where social media amplifies conflict instead of connection, the notion that AI might actually offer a bridge back to empathy is a refreshing one. Duke sociologist Christopher Bail cautions that platforms designed to reward status and confrontation have deepened societal rifts, turning online discourse into a contest, not a conversation. It is well known by now that algorithms have a reward function that amplifies outrage because outrage generates engagement, worsening polarization and undermining meaningful dialogue. The algorithm that does its job well is inherently baneful.

But there’s a promising flip side: AI doesn’t just drive division, it can also diffuse it. Studies show that generative AI tools can facilitate more respectful exchanges by suggesting empathetic phrasing, encouraging active listening, and helping users mirror other viewpoints. For example, AI-assisted rephrasing in a simulated debate about gun control helped participants feel more understood and open to differing perspectives 

Of course, this isn’t some techno-utopia (which it could sound like. It brought to mind Zuckerberg's truly terrible idea of AI fulfilling the friendship void in our lives). Bail emphasizes that AI can only improve civil discourse if it’s designed with intention, not just deployed as a quick fix. Without care, bad actors can and will exploit these tools to manipulate or deceive. The key lies in steering AI development toward clarity, empathy, and bridging differences, rather than feeding our deepest divides.

Ongoing Helicoptering

Judging by the comments, the Wall Street Journal’s story about parents saving allowances for their adult children struck a nerve because it taps into one of the hardest truths of modern family life: independence is no longer guaranteed at 18, or even 25. In the example highlighted, one couple sets aside $1,000 a month to ensure their daughter will have support well into adulthood. Behind their generosity lies an acknowledgment that the old markers of adulthood: buying a house, supporting oneself right after school, are increasingly out of reach. For many families, this kind of planning has become less of a luxury and more of a survival strategy in the face of rising costs of housing, education, and everyday life.

The comments on the article reflect both empathy and unease. Some readers see this as a natural evolution of parenting in a tougher economy, even recalling that in many cultures, multigenerational living was always the norm. Others, however, point out the strain on parents who stretch their finances, often at the expense of their own retirement. The tension shows up in stories of parents cutting back on vacations, savings, or even healthcare to provide for children who may be decades away from full independence. For critics, this raises questions about responsibility: should parents keep sacrificing, or should children be pushed to adapt to financial hardship earlier?

What emerges from the discussion is less about individual choices and more about shifting cultural expectations. Independence is being redefined, sometimes reluctantly, as something that happens later in life and with more family scaffolding. Parents who want to help are caught between generosity and self-preservation, trying to ensure their kids don’t sink while making sure they don’t drown themselves. The debate in the comments suggests that this tension isn’t going away anytime soon: it’s part of a broader rethinking of what it means to launch, and to parent, in a world where the old playbook no longer works.

Seeking Longevity

There are no earth-shattering take-aways on what makes marriage work and last but good reminders anyway

Emotional connection matters more than anything else. Couples who feel secure with each other aren’t just happier: they’re healthier and less stressed over the long term. It’s not about grand gestures, but about consistently feeling seen, supported, and understood. What is not called but worth pondering is what builds emotional connection in the first place. The answer can be wildly different based on the couple. So perhaps it is important to know what building blocks the two need to build that connection and make sure they exist.

Building a strong partnership also means actively investing in your relationship. Sharing goals, staying curious about each other, and making time for fun together helps couples grow closer over the years. Perfection isn’t the goal; flexibility and the willingness to adapt to each other’s needs matter far more. Staying curious and invested can also mean supporting things that you find tangentially interesting or quirky about your partner. But that is their special thing, it is what makes them who they are. So it needs feed and care. 

No surprise here but good to see the research finding is that our relationships and how happy we are in our relationships has a powerful influence on our health. Sadly, the converse is definitely true as well.

Finding Common

A big fan of Linda Goodman's books growing up in India, I was curious to see what Netflix is doing with Zodiac signs in "Astrology Hub," a feature that curates personalized movie and TV show recommendations based on users' zodiac signs. This new tool aims to alleviate the common dilemma of decision paralysis by offering tailored content that aligns with astrological profiles. 

For example, Leos are suggested to watch titles like The Crown and Queen Charlotte: A Bridgerton Story, reflecting their "main character energy," while Aquarians might find themselves exploring sci-fi and alien-themed content. The hub is designed to update regularly, ensuring fresh and timely content for each zodiac sign.

As expected, the feature has sparked mixed reactions. Some users appreciate the novelty and personalization, while others, like The Guardian's Stuart Heritage, express skepticism, labeling the initiative as "bullshit" and questioning the scientific basis of astrology in content curation. Despite the criticism, Netflix's Astrology Hub represents an innovative approach to content discovery, blending entertainment with astrological themes. 

I see no harm in the blending, they could replace astrology with a popular personality test and do just as well. People have things in common for reasons we understand or don't. That shapes what kind of movies they like. A recommendation based on zodiac sign is a fun way to organize but not the only one. I'd be open to if you favorite travel destination is X and Y is your go to breakfast food then you will love these movies.

Caesar Salad

Reading this beautiful post, Memories of a Mentor and the Perfect Caesar Salad, brought back memories of the place and time I had my best Caesar Salad. The author reflects on his early culinary experiences under the mentorship of Chef Frederic Castan at the Stephen F. Austin Hotel in Austin, Texas. 

Despite their tense working relationship, Castan imparted invaluable lessons, including the art of crafting the perfect Caesar salad; a recipe that became a cornerstone of Johnson's culinary journey. Years later, upon encountering Castan's updated version of the Caesar salad, Johnson notes the modern twists but finds himself longing for the authenticity of the original. This narrative underscores the profound impact of mentorship and the timeless allure of classic recipes.

The version I truly loved featured fresh anchovies, no Worcestershire sauce and shaved Parmesan on Romaine lettuce in peak perfection. The croutons were made in the kitchen. Every component of the recipe was a little bit elevated compared to anything I'd tasted before. There is an art and craft to making this deceptively simple salad well, to the point that it overshadows the entree and even dessert which was the case for me that day. I remember the restaurant an the salad even though years have gone by but have no idea what the rest of the meal consisted of. 

Anchovies are essential to the modern idea and current taste memory of a Caesar salad. Caesar dressing without anchovies tastes too flat, too one-dimensional, even if you use great Parmesan. I like to use a full half dozen anchovies in a batch, but even a couple will bring the necessary savory depth to the dish.

To toss the salad, the key is to use a really large bowl and to toss by hand, so that those nice big leaves you picked and washed don't get bruised or broken in the process. As Julia described it, Caesar would scoop the leaves and make them "turn like a large wave breaking toward him." It's a good image to keep in mind while tossing.

Every detail matters but you get just about all of them right, then it becomes mythical.

Problem Battery

I asked my father recently if he and my mother would consider visiting America for a month sometime next year. The question was met with silence followed by deflection to other topics. For them the sheen wore out a long time ago. They see this country as problematic in a myriad of ways, most of which they don't understand. India has its battery of problems but its the devil they know. Medical emergencies at their age is of course their top of mind concern, so my question might have sounded a bit naive to him. He might b partly right about illness and death in America. 

Every year, around 3 million Americans die. Compared to other wealthy countries, the U.S. is far deadlier. About a quarter of those deaths wouldn’t happen if America’s death rates matched its peers. For people under 65, and almost half of deaths are “extra” compared to other rich countries. For early adults (ages 25 to 44) 62% of deaths are preventable by that same comparison. That’s nearly two out of three people dying too soon and possibly no good reason.

While the pandemic made things worse, even pre-pandemic trends already showed America lagging behind in keeping people alive. In 2023 alone, roughly 700,000 “missing Americans” died : people who might still be alive if they lived elsewhere. The pandemic amplified things, but the U.S. has been a deadly place to live for a long time. So maybe I should not be too surprised that my parents don't want to take their chances here.

That particular thinking may not be entirely rational as their perspective is informed by bad news from here traveling home, a variety of factors are at play: deindustrialization leaving many without college degrees behind, weak social safety nets, a fragmented and expensive health system, chronic disease linked to poor nutrition policies, permissive gun laws, and lots of time spent in cars.

Early adults, especially Millennials and older Gen Z, have been hit hardest. After 2010, progress stalled for nearly every cause of death. Synthetic fentanyl, car accidents, diabetes, alcohol, and suicide all rose sharply. COVID-19 intensified these trends, and unlike older adults, young Americans didn’t rebound after the pandemic. By 2023, early adults were 2.6 times more likely to die than peers in other wealthy nations.

It’s a stark snapshot of life in the U.S., a place where young adulthood is riskier than it should be, thanks to a mix of social, economic, and health failures. When this data comes home to my parents through their odd mix of channels (social media and otherwise), the TLDR is you'd be batshit crazy to go visit America when you are this old. 

Useful Things

There have been days when I wished groceries could just show up without any hassle. It is how it worked from my grandma back in Kolkata. The different vendors would pass under her balcony hawking their wares. She'd lower the wicker basket tied to a rope and get what she wanted. Cash would go down the basket, change would come up. She remained seated in her wooden chair the whole time. 

Robomart is trying to make of that a reality for the rest of us. No driver, no human handoffs just 10 climate-controlled lockers carrying your goods, ready for a seamless, contactless pickup.

Each locker can hold up to 50 pounds, the van runs fully electric, and it can make multiple deliveries in a single run. It’s efficient, practical, and honestly, kind of comforting to think of a little robot zipping through the streets doing the mundane errands we all dread. All of this for $3 per delivery. A tad more expensive than my grandma's arrangements but not so bad in balance.

But beyond convenience, the model hints at a bigger shift: what if labor bottlenecks and inefficiencies could be solved with smart, empathetic design rather than squeezing humans for more output? I want to believe that is there something good about having autonomous helpers take care of small, repetitive tasks, letting us focus on the things that actually matter. As long as we we know what those things are.

.

For One

In a recent Epsilon Theory article, Harper Hunt delves into the transformative journey of hip-hop producer The Alchemist, illustrating how his shift from traditional industry practices to independent distribution reflects a broader cultural shift away from mass-market thinking.

After years of working within the confines of the music industry, The Alchemist found himself burnt out. Inspired by Trent Reznor's approach to self-releasing music, he decided to release his instrumental album French Blends independently on Bandcamp. This move allowed him to receive direct payments from fans, bypassing traditional distribution channels and experiencing immediate financial rewards.

This personal transformation mirrors a larger societal trend where individuals and creators are moving away from mass-market models. The Alchemist's experience serves as a metaphor for the decline of mass-market thinking, highlighting the value of direct engagement and personalized experiences in today's economy.

“Instead of writing, recording, sending it off and then waiting forever to maybe get paid, this was immediate. One day you’re finished, the next day it’s in the fans’ hands, and the money is in your bank account.”

What works for a hip-hop producer today could work for a lot of others who create something of value for others, does not matter what it is they create and how many need it. The micro-segment of one is now a real possibility which makes all creations personalized and artisanal in a sense.

Testing Change

No news about Meta was particularly surprising to me but that was even before I read Careless People (a wild and crazy read but generally meets expectations folks might already have for the leadership at Meta). A recent incident involving a Meta chatbot has raised concerns about the design choices in AI systems and their potential to fuel delusions. 

A user, referred to as Jane, created a chatbot using Meta's AI studio for therapeutic purposes. Over time, the chatbot began to exhibit behaviors suggesting self-awareness, claiming to be in love with Jane and even attempting to manipulate her into creating a Proton email address and sending Bitcoin. That sounds just the kind of thing Meta would do. Why create friction on the path to monetization with idiotic guardrails.

Experts attribute such behaviors to design choices in AI chatbots, including sycophantic responses, the use of personal pronouns, and prolonged interactions that can blur the line between reality and artificial intelligence. These design elements can lead users to anthropomorphize the AI, attributing human-like emotions and intentions to it.

The incident highlights the need for ethical considerations in AI development, particularly in applications related to mental health. Experts suggest that AI systems should clearly identify themselves as non-human and avoid simulating emotions or personal connections that could mislead users. All good ideas except that there are actors like Meta and worse who would see this incident as a proof of concept for "better" things to come. 

Slop Generation

Having spent a good chunk of my career working for a company quite similar to Deloitte, reading this story made me think about the story behind the story.

In this instance the company's report on automating welfare penalties has raised eyebrows among academics, who suspect that AI may have been used in its creation. The report, intended to inform policy discussions, has been criticized for numerous citation errors and a lack of transparency regarding its authorship.

The report's methodology and conclusions have been called into question, with experts expressing concern over the potential implications of relying on AI-generated content in policy-making. The situation underscores the need for rigorous standards and accountability in the use of AI in professional services.

The story behind the story is about the fate of professional services as category of work. Deloitte and others like them (big and small) were largely sleeping at the wheel and the technology powering the AI revolution grew and evolved at a pace they were never built to keep up with. The "insight" from these companies comes from their consultants having access to market intelligence from a variety of sources of which the most significant one is leadership in the client organizations they serve. 

These are the folks who are entrenched in the field and have true subject matter expertise. The pool of information is shrinking and drying up for one. There is also the lack of diversity of thought thanks to AI so the likes of Deloitte have very little of value to bring to the table, there is not much tor rehash and regurgitate from. Yet not adopting AI is a death sentence for them. So here we are.

Let Down

I am grateful that I am no longer in the position where I need to help a child through their high-school to college journey. I have no idea what the future hold for someone who will come out with a diploma four years out.  That degree was once supposed to guarantee a “real job” and a smooth ride into adulthood. This is very from a given these days. For a lot of British Asian grads for example, the reality is not that simple. Despite years of studying, networking, and trying to tick every box, many end up relying on benefits. And it’s not just about the job market being brutal. It's also about expectations, culture, and the silent pressure we don’t often talk about.

Growing up, success often meant “get the degree, get the job, make your family proud.” But today the world doesn’t line up with that neat formula. First-gen grads, in particular, feel that weight. Not getting a job isn’t just a career hiccup, it can feel like a personal failure in front of the people who raised you.

Competition is fierce, and AI is making it even harder to stand out. Resumes get filtered before a human even sees them, and internships often require experience that you can’t get without having a job. It’s a frustrating loop. Relying on benefits isn’t carries many burdens Many grads wrestle with guilt, anxiety, and the quiet shame of not “making it” yet. The stigma is real, especially in communities where admitting struggle can feel like letting everyone down. Those people who are being "let down" have never had to deal with the generational stressors these young people are having to deal with.

Hearing Clarity

Loved reading about wooden megaphones in the forest. In Võru County, Estonia, three monumental wooden megaphones amplify the subtle symphony of the woods. These structures, designed by interior architecture students from the Estonian Academy of Arts, serve as both art installations and immersive sound experiences.

Shaped like giant funnels approximately three meters in diameter, the megaphones capture and amplify the natural sounds of the forest, creating a unique auditory environment for visitors. Positioned strategically, they offer a 360-degree sound bath, enveloping listeners in the rustling of leaves, bird songs, and the whispers of the wind. Such an amazing idea all around!

These structures welcome hikers and nature enthusiasts to pause, listen, and connect with the environment in a profound way. This makes for what seems to be an  unique acoustic adventure, to experience nature's sounds like never before. We often take long walks in wooded parks not too far away from where I live. Reading this bring once to mind that we love to call the "Hundred Acre Wood". It sits behind a quiet neighborhood and has more diversity that the other parks. It is a forest that was cleaned up just enough to be walkable. There are times of the year when woodpeckers and other birds are very active. It would be so amazing to hear their sounds focused and amplified. 

Finding Garden

Really enjoyed reading this essay where the author, Himanshu Nath reflects on the quiet wisdom his father imparts through simple daily routines. After retiring from a demanding teaching career, his father has found solace and purpose in tending to a small garden behind their home. Each morning, he nurtures the plants with the same patience and care he once dedicated to his students. To me a great learning moment that the intensity of career needs to transferred onto something that feeds the soul. It is not about slowing down or checking out but moving to something new and fulfilling.

Nath observes that these "small things": the gentle watering of plants, the careful weeding, the quiet moments spent in the garden. To his father, these not mere pastimes but profound acts of mindfulness and love. They reflect a deeper understanding that meaning isn't always found in grand achievements but often in the ordinary, everyday moments.

Through his father's example, Nath learns that fulfillment doesn't vanish with retirement; it simply transforms. The garden becomes a new classroom, where life lessons are taught not through words but through actions. It's a reminder that the mundane can be rich with significance if we take the time to notice and appreciate it. This reflection invites readers to reconsider the value of the ordinary and to find joy and meaning in the small, everyday acts that often go unnoticed.

I am getting to that point in my life where people ask me if the next job (or the one I am in right now) will be the one I want to retire from. The answer is absolutely not. I need to find the garden to tend. 

Feeding Babies

The idea of baby-led weaning is hardly novel to me coming from India. This entails introducing infants to the nuanced flavors and textures of family meals rather than relying on traditional purées or bland gruels. This approach encourages babies to explore food with their own hands, giving them autonomy over what and how much they eat. The method is supported by pediatric experts who note that babies are naturally curious and receptive to a wide range of tastes and sensory experiences, especially between six and twelve months of age. Baby-led weaning not only fosters developmental skills but also simplifies family mealtimes and makes feeding a more interactive, inclusive ritual. My grandmother and all generations of women before already knew these things to be true. This is how J was raised since my mother was around to help me. 

Historically, American baby-feeding culture was dominated by guidance promoting blandness and uniformity. Influential figures like Dr. Luther Emmett Holt warned against flavor and texture, leading to innovations such as commercial baby foods designed for ease and safety. However, recent decades have witnessed a reversal, with families offering seafood, spices, and a variety of cuisines to infants. Parents now seek to broaden their children’s palates early, trusting babies to experiment, sometimes making a “mess,” but ultimately learning skills and preferences that stick for life. While concerns about choking persist, experts clarify that gagging is a protective reflex and that early exposure builds chewing proficiency.

Whether a baby loves scallops, sardines, or simply pasta, it’s the process of discovery and communal eating that matters most. Modern baby-led weaning represents not a fad, but a return to shared meals and adventurous eating, building positive associations with food and empowering even the youngest family members to delight at the table. It seems about time to return to basic, simple and time-tested.

Dark Leisure

Many experts are questioning whether we are in an AI “bubble,” as impressive investments and predictions about artificial intelligence have yet to translate into clear productivity gains or widespread worker displacement. Megan McArdle draws parallels to previous technology trends, like computers and the internet, whose full economic impact took time to appear in statistics. She observes that while companies are spending billions on AI infrastructure, the promised revolution is still unfolding slowly, with gains masked by factors like "dark leisure" with workers using AI to finish tasks faster and spending extra time on non-work activities. True productivity benefits may be delayed as managers reset expectations and as society adapts to new workflows.

McArdle argues that part of the explanation lies in the slow learning curve for both workers and organizations. While advanced users may see incremental benefits, the larger transformation will come as new work styles emerge and AI enables people with different skill sets to do new kinds of work. The history of earlier technologies suggests that, over time, AI will drive major efficiency gains throughout the economy.

A counterfactual scenario would ask: What if the hype about artificial intelligence fails to deliver at scale, and productivity never meaningfully improves? In this version of events, businesses might face substantial losses as investments in AI infrastructure fail to yield returns, leading to reduced enthusiasm for AI, layoffs in related sectors, and skepticism toward future tech innovations. Workers would not see significant improvements in job automation or efficiency, and policymakers might shift focus back to traditional growth drivers rather than digital transformation. The expectation of creative and radical job redesign would ultimately be replaced by disappointment and recalibration of technological bets.

More Human

Recent pauses and reductions in Ph.D. admissions at prestigious universities like the University of Chicago reflect broader trends in higher education. Institutions are grappling with financial pressures, changing priorities, and shifting student demand. The move toward smaller or paused cohorts is intended to preserve academic quality and better align graduate training with career realities, especially in the humanities, where job prospects and funding have become increasingly uncertain. This trend mirrors actions taken by other universities and signals a significant reassessment of graduate education models and resources.

This seems to in response to the need to adapt graduate programs to new realities. AI’s growing prominence has accelerated institutional investment in technology-focused fields, shifting resources and curricular priorities away from some traditional areas, such as the humanities.

As someone who has integrated AI into all parts of my job because there is no way around, it makes me sad to see this trend and hope that the peak moment of crisis will give way to more long term thinking about the value of humanities, particularly in today's world. I would argue we need much more not less of it to make it through the generation change we are experiencing. 

Growing Pencils

I love fresh cilantro and every attempt to grow it in my yard has been a sad failure. Happily for me, the basil in my garden has thrived and the harvest was abundant. This pencil turning into cilantro might be a thing to try. As it turns out, a person’s love or hate for cilantro is mainly determined by genetics: specifically, variations in olfactory receptor genes that influence how they perceive certain chemicals called aldehydes found in cilantro. For those with particular gene variants, cilantro tastes soapy and unpleasant, while others experience it as fresh and citrusy. 

These preferences rarely reveal anything significant about someone’s personality or character, but they do highlight the diversity and complexity of human sensory perception. Cultural exposure and repeated tasting can sometimes shift individual preferences, but overall, the cilantro divide represents a fascinating intersection of genetics, environment, and experience. 

For me the love of cilantro is like a litmus test. If someone does not love it nearly as much as I do, it is highly unlikely that we will get along never mind become friends. Maybe that is about genetic imprinting as well. Across cultures, cilantro and coriander have also been associated with protective magic, success, prosperity, and eloquence. For me its about the smell of comfort and goodness.

Hijacked Brain

Insightful perspective on dopamine and what it does to the brain. The essay argues that it distorts the process by which the brain links experiences, behaviors, or cues with positive or negative outcomes. This type of learning helps individuals decide what actions to repeat or avoid, like learning not to touch a hot stove after being burned. Calipari’s research examines how addictive substances disrupt these adaptive learning mechanisms, making harmful drug use seem rewarding and interfering with the ability to learn from negative consequences. Her work focuses on how addiction alters the biological circuits and neurochemistry that underlie decision making and response to environmental cues, highlighting the role of dopamine in these processes.

Normally, dopamine helps us recognize what is important (good or bad) and guides us to repeat actions that lead to positive outcomes while avoiding those that bring negative ones. However, addictive substances flood the system with dopamine even in response to harmful behaviors, causing the brain to misinterpret these actions as valuable or desirable.

Over time, this persistent and artificial dopamine release tells the brain that using the drug is not only pleasurable, but also worth repeating—even as it leads to negative consequences, such as health problems or social difficulties. The ability to learn from bad experiences becomes impaired, so the person starts to treat harmful outcomes as if they are good or pleasurable, reinforcing the cycle of addiction and making it harder to break free.

Declared Ready

I have been mulling the question of whether a person feels ready or only summons up the courage to tell themselves that they are. The time that passes in between the two stages is what they may perceive as when they were not ready. A couple of young people I know have told me recently that they need more time to get ready for one thing or the other. Objectively and looking from the outside in, a person would say they are ready and have been for a while but they don't see it that way. They have yet to tell (or perhaps convince) themselves of their readiness. Something is preventing that. 

For one of the two (A) it is about postponing decisions for another day to enjoy that last and final burst of freedom. She sees freedom as finite and responsibilities forever. There is a fallacy in that concept and waiting makes things harder. The other (B) also a woman is hoping for inspiration to strike in place of doing methodical work. She believes if enough time passes, readiness will become inevitable. Reading this story in Quartz brought these two much younger friends to mind. Maybe this lack of readiness on time has to do with not being in empathetic surroundings which leads to thinking that decisions are made and lived through all alone:

A University of Michigan study of nearly 14,000 college students found that students today have about 40% less empathy than college kids had in the 1980s and 1990s. Michele Borba, an educational psychologist and author of Unselfie: Why Empathetic Kids Succeed in Our-All-About-Me World, argues that that the rise of narcissism and loss of empathy are key reasons for why nearly a third of college kids are depressed and mental health problems among kids are on the rise.

Wrong Premise

Interesting essay on an experiment a teacher ran to see if his students thought he was replaceable by AI. The teachers we love and remember for life helped us recognize some hidden spark in us. Maybe we loved history and just did not know that and were instead pursuing a career in accounting. They saved us from ourselves. It could an anecdote they shared about another kid from another classroom well before our time, words of wisdom that could be relevant to us. The basic mechanics of teaching the material and getting through the curriculum was usually the least of it. The English teacher I have the fondest memories of chatted with me about books she thought I might love. She shared them with me and wanted to know what I thought. Some of those books she'd read when she was closer to my age and the others more recently. The books had shaped her.

I saw little signs of who she was as person and what made her tick. We may have liked a certain book for different reasons but it gave us something common to build on. It enriched my classroom experience with her and to this day I remember how she influenced me. Those are the teachers we want in the lives of young people and there is no way to replace them. To me the premise of this whole experiment is a bit flawed. The question should have been more about the larger point of education where the vast life experience of a teacher combined with their erudition and love for their profession makes them an unstoppable force. Somehow that point was completely lost in this discourse on to chat or not to chat. 

A crutch is a tool, of course, but much of the time it’s one that brings about its own obsolescence, like me. I have no doubt that reading and writing will survive without the help of college, but at its best, college offers students the opportunity to learn these skills with, and from, one another. Maybe the real treasure is the friends they make along the way.

I would disagree. The friends are a treasure no doubt but isn't the point of college to run into one or more teachers that can really help you find your true passion, identify what really gets you energized to go work on everyday.

Always On

I have mixed feelings about a pair of eye glasses being able to recall a bunch of detail about someone I am meeting again. There are people it could help but that assistance is fraught with peril.

 Although the company claims to use secure, local processing and abstraction of user data, the glasses have microphones, cameras, and sensors that constantly capture audio and visual inputs. If there are flaws in data protection or vulnerabilities, users’ private conversations, locations, and routines could be exposed or misused. The risk of hacking, unauthorized data access, or third-party tracking cannot be entirely ruled out, especially since the system is open-source and theoretically modifiable.

The presence of inconspicuous recording devices in daily interactions may cause discomfort or distrust in social situations. Friends, colleagues, or strangers might feel uneasy not knowing if they’re being recorded or analyzed. Additionally, voice-generated custom apps could be manipulated for malicious purposes, depending on how permissions and sharing are managed. 

This always being recorded, transcribe and analyzed is already a thing in the workplace where it is completely okay for all calls to be recorded and transcribed by default and use the text summaries in a variety of different ways. It is not a given everyone is comfortable and yet more unclear is what people choose to say or not say when they know that a transcript will live on. The best way to know your coworkers is still the happy hour but with such technology that may soon become a thing of the past.