mintconditioned

The sun peered nosily through the windows of the Manhattan federal courthouse.

“The moon will have a hoot when it hears this,” it thought.

Judge P. Kevin Castel squinted at the citations, gleaming Times New Roman ghosts, and frowned. The sun leaned in, delighted, as the judge opened his mouth and asked a question that would spark a hunt for the ages:

“Are these citations real?”

The answer was, unfortunately, no.

The sun was expressionless for a beat. Then an uneasy yet knowing smile spread across its face. Humans had always been fond of shortcuts. It had watched them invent wheels, elevators, calculators, and remote controls. If there was a faster route from Point A to Point B, someone would eventually take it.

Attorney Steven A. Schwartz was no exception to this principle. Months earlier, while researching for this case, he had found a seemingly innocent loophole to his late-night battles with case law: ChatGPT. Its promise was irresistible: ask a question, get an answer. No dusty archives. No endless searches through legal databases. Just answers.

The citations ChatGPT produced wore the costume of expertise perfectly. They had case names, docket numbers, quotations, and judicial language. Yet when Judge Castel went looking for them, they were nowhere to be found. The cases existed in the same place as dragons, Atlantis, and every group project where everyone claimed to do their share.

What followed became one of the first great cautionary tales of the AI era.

The problem was not that the machine made a mistake. The sun had watched people confidently insist that the Earth was the center of the universe for centuries. Making mistakes was hardly new.

What was new was how persuasive the mistake was.

The fabricated cases were dressed in intelligence’s robes of authority. They sounded intelligent, they looked intelligent, and most importantly, they saved time.

And saved time, the sun had noticed, was often humanity's favorite argument.

For generations, expertise required effort. Knowledge came with hours spent poring over textbooks, researching dead ends, and verifying with mentors. The process itself, though often irritating, was quite useful. A person who spent days working through a problem usually understood something about it by the end.

Artificial intelligence rearranges that relationship. Scholars call this rearrangement epistemic compression: the collapsing of a lengthy reasoning process into an immediate answer. Think of it as getting to the destination without necessarily taking the journey.

The sun found this unsettling. For thousands of years, human beings had complained about how difficult thinking was, and now they had invented a machine that agreed. Why did it feel uneasy about this advancement?

The sun was not vain in its thinking, however. The consequences of epistemic compression are visible in everyday life. The University of California updated its guidelines in 2023 to allow the use of generative AI in academic work, but requiring students to verify information and disclose usage. Other schools, including selective New York City high school, Stuyvesant High School, have also issued guidance encouraging transparency around AI-assisted work.

Additionally, law firms such as Latham & Watkins have warned attorneys that AI-generated citations and references must be independently verified before being submitted in any legal filing and likewise, consulting firms like McKinsey emphasize that generative tools can help with drafting, but cannot replace fact-checking.

Ted Chiang anticipated a similar shift in his short story Catching Crumbs from the Table. He suggests that as intelligence-enhancing tools become widely available, ideas and skills that once distinguished a few may become more commonplace. In other words, when advanced tools are accessible to everyone, individual achievements may appear less extraordinary, not because people are less capable, but because the tools level the playing field.

Think of it like Willy Wonka’s special golden ticket. What made it a press sensation was that only five kids, Augustus Gloop, Mike Teavee, Charlie Bucket, Veruca Salt and Violet Beauregarde, received one. Now, if everyone had received one, the golden ticket would be much less frenzied over and more like an “oh, nice.”

Artificial intelligence intensifies this uncertainty. Due to its generative nature, artificial intelligence does not simply assist thinking, but rather recontextualizes the conditions under which thinking happens.

In the past, when a question was asked, there was a significant period of waiting before an answer was presented. In this time, one may go for a walk, play with their pet, or, if they were nerdy like that, doubt and refine their ideas to form a consistent, strong conclusion.

Artificial intelligence shortens that waiting period dramatically. The same technology that can summarize hundreds of pages in seconds can also produce convincing errors with equal speed. In the case of Schwartz, fake legal citations moved from generation to submission before anyone stopped to verify them. The issue was not that the information was wrong, but that it arrived so quickly and confidently that skepticism never had a chance to catch up.

In this way, some forms of deep, independent thinking may become less necessary in day-to-day life. Generative systems do not prevent people from thinking carefully, but they can reduce the pressure to do so for routine tasks. Over time, skills that are exercised less may become less sharp if not intentionally practiced.

Yet there were other considerations. The sun had watched humans declare the calculators as witchcraft, yet they did not destroy mathematics. Likewise, search engines did not destroy knowledge. In the same way, AI, people argue, may free individuals from mundane mental labor and let them focus on higher-order thinking.

Yet this optimism rests on an important assumption: that the skills being outsourced can be safely neglected.

Unfortunately, understanding does not work that way.

The ability to evaluate an argument depends on prior knowledge. The ability to ask good questions depends on familiarity with a subject. Judgment grows from repeated encounters with confusion, error, uncertainty, and effort. This process produces questions that lead to innovation and advancement in creative and academic spaces. However, when enough of those experiences are removed, the foundations begin to weaken.

The lawyers in Mata v. Avianca did not lack access to information. If anything, they had too much. What they lacked was something less visible: the habit of doubt. When the machine produced a clear, confident answer, the users accepted it without any cross-checking or further research. The truth is that for many users, confidence usually overrides correctness.

As artificial intelligence spreads through schools, workplaces, and daily life, societies face a difficult question. If knowledge can be generated without understanding, how should competence be measured? If polished work no longer proves expertise, what does?

Institutions might rely more on oral examinations, live problem-solving, and demonstrations of reasoning. Or they may accept a future where it is often impossible to know whether intelligence belongs to a person or a machine speaking through them.

The challenge, then, is not that artificial intelligence will make humans less intelligent. Rather, it may reduce the everyday need to rely on certain kinds of thinking, which could change how skills are developed and applied over time.

The gavel rang clear in the sun’s ears as the Court dismissed the case. It could almost feel the $5,000 fine weighing down on Schwartz’s shoulders. As the legal analysis was described as “gibberish,” the sun drifted across the courthouse floor and smirked at the irony. Humanity had spent centuries building machines to save time and effort, and now it had built one capable of saving thought itself.

“The moon is never going to believe this,” it thought.

Works Cited

Bhatnagar, Pranav. “Trust Amplification Exploits in Human to AI Security Teams.” Zenodo (CERN European Organization for Nuclear Research), 29 Jan. 2026, doi.org. Accessed 25 Apr. 2026.

Bjork, Robert A., and Elizabeth Ligon Bjork. “Desirable Difficulties in Theory and Practice.” Journal of Applied Research in Memory and Cognition, vol. 9, no. 4, Dec. 2020, p. 475. ScienceDirect, https://doi.org/10.1016/j.jarmac.2020.09.003.

Drijvers, Paul, et al. Use of Technology in Secondary Mathematics. Research Report, International Baccalaureate, 2014, p. 22. HAL, hal.science/hal-01546747.

Goh, Ethan, et al. “GPT-4 Assistance for Improvement of Physician Performance on Patient Care Tasks: A Randomized Controlled Trial.” Nature Medicine, vol. 31, no. 4, 5 Feb. 2025, p. 1233, https://doi.org/10.1038/s41591-024-03456-y.

Hogan, Alexander, et al. “On the Cruelty of Computational Reasoning.” Anaesthesia, vol. 63, no. 1, 2008, p. 4. Wiley Online Library, doi.org.

Jose, Binny, et al. “Epistemic Authority and Generative AI in Learning Spaces: Rethinking Knowledge in the Algorithmic Age.” Frontiers in Education, vol. 10, 19 Aug. 2025, doi.org.

Kabashkin, Igor. “Cognitive Atrophy Paradox of AI–Human Interaction: From Cognitive Growth and Atrophy to Balance.” Information, vol. 16, no. 11, 2025, p. 1009, https://doi.org/10.3390/info16111009.

Klein, Colin R., and Ronald Klein. “The Extended Hollowed Mind: Why Foundational Knowledge Is Indispensable in the Age of AI.” Frontiers in Artificial Intelligence, vol. 8, 2025, doi.org.

Matta, Daoud. “The Present and the Next: A Markovian Perspective on Epistemic Compression in AI and Human Cognition.” Zenodo, 7 Apr. 2026, https://doi.org/10.5281/zenodo.19447042.

Miele, David B., and Daniel C. Molden. “Naive Theories of Intelligence and the Role of Processing Fluency in Perceived Comprehension.” Journal of Experimental Psychology: General, vol. 139, no. 3, 2010, p. 535. APA PsycArticles, https://doi.org/10.1037/a0019745.

“New College Board Research: Faculty Express Near-Universal Concern That Student AI Use Undermines Original Writing and Critical Thinking.” Newsroom, College Board, 25 Feb. 2026, https://newsroom.collegeboard.org/new-college-board-research-faculty-express-near-universal-concern-student-ai-use-undermines.

Oakley, Barbara, et al. “The Memory Paradox: Why Our Brains Need Knowledge in an Age of AI.” The Artificial Intelligence Revolution: Challenges and Opportunities, Springer Nature, 2025, p. 39.

Restrepo, Dylan J., et al. "When AI Output Tips to Bad but Nobody Notices: Legal Implications of AI's Mistakes." arXiv, 25 Mar. 2026, doi.org.

Revell, Timothy, et al. “ChatGPT versus Human Essayists: An Exploration of the Impact of Artificial Intelligence for Authorship and Academic Integrity in the Humanities.” International Journal for Educational Integrity, vol. 20, no. 1, 21 Oct. 2024, p. 18, https://doi.org/10.1007/s40979-024-00161-8.

So if you saw my last post, I posted “clean girl is over” and thought I was just spiteful because I look terrible in a slickback.

But I did a lot of research since then.

A lot.

So I just wanted to share what I found since I promised to update you guys.

You guys gave me more trends to test, so I expanded to these aesthetics:

Balletcore

Office siren

Indie kid

Tumblr grunge

Dark academia 

Y2K 

Coquette

Shoujo girl

VSCO girl

Downtown girl

Acubi

Goblincore

I basically tried to measure waste from aesthetic trends without directly tracking landfill data (which is almost impossible to get in real time).

So I used three proxies:

1. Value collapse

I measured this as % of listings under $15

Because in resale, under $15 = near-discard tier

2. Listing surplus

This is just total resale listings

More listings = more material committed to the trend

Cottagecore is insane here btw (356K listings)

3. Supply-demand mismatch timing

Basically a trend peaks, demand falls, but supply doesn’t

The longer prices stay low after the peak, the more likely that stuff never gets used again

Then I combined all three into one number:

Trend Waste Index (TWI)

It weights:

– cheapness (0.4)

– volume (0.4)

– how long the oversupply lingers (0.2)

And the ranking is kind of brutal:

Which leads into the second thing I measured:

Trend Half-Life

This is just:

How long until a trend loses half its attention?

Average across everything: ~2 years

But here’s a pattern I’m noticing:

Older trends (2020-2021):

~2-3 year EHL

Mid trends (2022–2023):

~1-2 EHL

New trends (2024-2026):

sometimes under 1 year

This shows trend decay is accelerating.

And when you line that up with resale data, you get this pattern:

Fast spike

Instant overproduction

Price collapse

Resale flood

Which explains why prices all cluster around like $11-$13

Everyone is undercutting each other because there’s too much supply

Different types of trends behave differently though:

Microtrends (office siren, acubi)

explode

die

create the most waste

Slower cultural trends (cottagecore, tumblr grunge)

decline gradually

still wasteful, but less abrupt

Revival trends (Y2K, dark academia)

never fully die

actually less waste because demand keeps coming back

This is basically a market failure. platforms and influencers generate demand while holding zero risk, brands produce fast and sell volume while offloading excess, and consumers and resellers are the ones stuck with depreciation, while the environment absorbs whatever is left.

The system overproduces by design because the decision-makers don’t bear the downside, which is what creates a negative externality. And it’s getting worse as everything speeds up. Trends move faster, production reacts instantly, economic life gets shorter, and waste per trend increases.

Resale markets don’t actually solve this; they just delay disposal while prices collapse. So trends aren’t just cultural cycles anymore; they’re compressed economic cycles, where value is extracted quickly at the top and waste is pushed downstream just as fast.

So I tested some other trends from Italy and South Korea. So far, they show me similar patterns, but I need more South Korean trends to really test my study. Five South Korean trends in total were used:

발레코어

하이틴

고프코어

Y2K

우아한

if you know of any more, please tell me bc i really am invested in this project.

also, sorry if any of this was unclear or anything; i’m writing this kinda late and my brain is fried. if you have any further questions feel free to message me and ask! if you want you guys can look more at my notes: https://drive.google.com/drive/folders/1DS2rFtq7b8t9jRk-l74gw8Ozb8qbaW34?usp=sharing they’re not finished but they give you more of an explanation than this post probably will lol.

So I really don’t know what exactly I did, but I was on Instagram and yo. Look, I wasn’t ever a fan of the "clean girl" trend, but I just found it fascinating how everyone’s ditching it for 2016 Zara Larsson coded makeup and outfits. Like good on you, but I was just thinking about what would happen to all the stuff from the clean girl era. I wanted to see whether certain aesthetics actually "die" or if we just stop talking about them. So I pulled Google Trends data. Then I checked resale markets. Then I made a spreadsheet.

I became slightly obsessed and turned it into a research question: How long does an aesthetic trend retains economic value after peak attention

If trends are accelerating, are they also collapsing faster? And if they are, what happens to all the stuff produced for them?

The Method

I picked three aesthetics:

Clean girl

E-girl

Cottagecore

For each one, I:

1. Mapped lifecycle phases using Google Trends

2. Identified peak attention

3. Took cross-sectional resale snapshots (30 items each, clothing only, no bundles, no joke prices)

4. Measured:

Median resale price

% under $15

Total listing volume

I treated resale value collapse as a proxy for economic devaluation.

Because once an item can't retain resale value, it's economically near-discard.

The Concept I Ended Up With

Economic Half-Life =

Months after peak attention until median resale price falls below $15.

Basically: how fast does a trend lose market survivability?

What I Found:

Cottagecore

Massive listing volume

73% under $15

Still huge resale presence years after peak

Long decay tail

E-girl

Peaked early (2021)

Lower listing volume

53% under $15

Slow bleed into low-value persistence

Smaller volume, but long persistence.

Clean girl

Slower rise (2022–2026)

Hit peak January 2026

83% under $15 almost immediately

Extremely tight price clustering (~$11)

This one collapsed economically almost instantly after peak attention.

Which suggests something new: Supply elasticity is now extremely high.

Production reacts so fast that by the time attention peaks, the market is already saturated.

I combined three proxies:

Share under $15 (value collapse)

Normalized listing volume (scale of material committed)

Post-peak persistence (how long low value lingers)

Cottagecore had the highest waste intensity

E-girl was persistent but smaller-scale waste

Clean girl had rapid short-term waste spike

The bigger pattern is that trend cycles are shortening.

Older trends take 12-18 months to collapse and ~1-3 years before resale value really died, but recent ones take 1-3 months to resale collapse and have under 6 months of survivability

Now because my entire page is about who profits and who holds risk, let’s analyze that real quick.

1. Platforms & influencers

Profit from attention velocity

Hold no inventory

2. Producers

Profit from volume

Offload surplus

3. Consumers

Absorb depreciation

Internalize resale failure

4. Secondary markets

Attempt to clear oversupply

Normalize ultra-low prices

Environmental systems absorb what markets can't, and economic devaluation precedes physical waste.

So what supreme is that trends behave like speculative assets. Early adopters extract symbolic and sometimes economic value, late adopters absorb depreciation, and platforms profit from volatility itself. As cycles shorten, the gap between profit extraction and waste generation widens. Costs move downward when move upward.

So what exactly I'm gonna do with this, I'm not sure. But I definitely want to expand because this is interesting, so here’s my very rough plan:

Expand to 10+ aesthetics

Build an Aesthetic Volatility Index

Stress-test the Economic Half-Life model

See whether post-2023 trends collapse nonlinearly faster

Anyway.

I’m hooked.

If you guys want, I'd really appreciate if you could recommend different aesthetics for me to do next, or tell me if you want a more in depth guide on how to do this yourself.

But anyway.

i am a lot of things. Today, I'm inspired.

There is a typical type of poem that places in competition. it'll have a clear metaphor, clear speaker, clear emotional arc, etc. Clarity is the key word here.

This poem is often well made. It is often sincere. It is often about loneliness, memory, grief, or becoming. It does not demand too much time. It does not require rereading. It does not make the reader feel lost, implicated, or uncertain of their footing.

It is, above all, legible.

Most competitions, workshops, and institutions prefer this legible writing style. They like poems which behave in neat ways and demonstrate skill without stepping out of the box and challenging the sheer markers of skill.

This is not a complaint about judges, but an observation of the systems in which they operate in.

I understand. Systems are neat, easy to categorize, and need rules to function. Rubrics require clarity. Time limits require efficiency. When hundreds or thousands of submissions pass through a small number of hands, work that can be “understood” quickly will always have an advantage over work that asks the reader to sit, to stumble, to doubt their own reading.

But the problem is, poetry, and writing in general, is not born from rules.

Poetry comes from pressure, and fracture, and language breaking under the weight of what it’s trying to hold.

The poems we now idolize were rarely respected in their own time. Shakespeare invented words because the existing ones were insufficient. Dickinson tore syntax apart. Whitman ignored meter entirely. None of them were writing with a rubric in mind. None of them were optimizing for clarity as it is currently defined.

They were not trying to place.

They were trying to say something that did not yet have permission to exist.

When we teach emerging writers that success looks like "sticking to the rules", we risk training them out of the very instincts that make their work alive. We ask them to write poems that will be rewarded, not poems that will cost them something to make.

And when those safer poems succeed, we call it merit.

Meanwhile, the work that is strange, volatile, structurally unstable, the sheer work that might be ahead of its time, or simply outside the current aesthetic bandwidth, often disappears quietly. Not because it is bad, but because it is "expensive" to read.

Rejection becomes especially cruel. Not just because a piece was turned away, but because the writer is left wondering whether the problem was the work or the fact that the work refused to behave.

A poem can be clear and still be empty.

A poem can be confusing and still be honest.

If we only reward the work that fits cleanly into our existing frameworks, we should not be surprised when the field begins to feel repetitive. We should not be surprised when innovation migrates into hybrid forms, into digital spaces, and into places that are not yet legible enough to institutionalize.

We live in a world that openly celebrates innovation as long as it happens in the “right” fields.

In STEM, innovation is practically currency. We praise new technologies, new systems, and new ways of thinking. We expect expansion. If someone proposed building the same computer we had twenty years ago, they would be laughed out of the room.

But in areas like writing, fashion, and other creative fields, innovation is treated with suspicion.

In fashion especially, trends repeat themselves in increasingly shorter cycles. Silhouettes, aesthetics, even entire subcultures are recycled, rebranded, and resold. In writing, the same emotional arcs, narrative structures, and aesthetic preferences circulate through workshops, journals, and competitions. The work is often competent. It is often polished. And it is often familiar.

And so, it will come to no surprise to you, that these fields are then dismissed as unserious luxuries, especially when compared to STEM careers that are framed as practical, future-facing, and necessary (and they are, but I am not getting into that whole debate right now.)

The issue is that we no longer reward innovation when it makes us uncomfortable.

We live in what is frequently called a “digital world,” as if that were a natural state rather than a choice. But this world was built. It was not inevitable.

When early computers emerged, people were afraid of them and questioned what this technology would do to attention, labor, and human connection. Expansion was slow, contested, and deliberate.

Now, in 2026, expansion is assumed. We move faster without asking whether speed is the same as progress. We accept constant updates as normal.

And yet, in creative fields, we resist expansion with remarkable force.

True innovation, whether technological or artistic, requires tolerance for confusion, and a willingness to sit inside unfinished ideas. STEM has institutional frameworks that protect this process: research grants, labs, long timelines, and an understanding that failure is part of discovery.

Art does not.

Instead, writers and designers are often asked to justify their work immediately. To 'explain, summarize and make it accessible.' Work that resists this demand is dismissed as self-indulgent, pretentious, or impractical.

We do not ask engineers to make their work emotionally relatable.

We do not ask scientists to ensure their discoveries are immediately comforting.

So why do we deny that same allowance to writing, fashion, and other creative disciplines, which are fields that shape how we understand ourselves, our bodies, and our relationships?

Expansion in writing does not mean abandoning meaning, and expansion in fashion does not mean rejecting wearability.

It means accepting that progress sometimes looks strange before it looks useful.

If we want art that feels alive again, we have to allow it the same generosity we extend to technology: time, patience, and the right to fail in public.

Innovation does not belong to STEM alone.

some images are not neutral, nor are they accidents. they're designed to provoke a reaction--shock, jealously, disgust, even defensiveness--because emotion drives engagement, and engagement drives profit.

when you see a room full of makeup, a 30-step skincare routine, a even just a "quick haul", the goal isn't to inform you, but to make you feel something. that feeling is usually insecurity or outrage, mainly because both keep you scrolling.

some people feel inferior and ask "why don't i have this? am i behind?" others feel morally superior and declare "this is disgusting. people are so wasteful." both reactions still boost the post. both still feed to algorithm. the key point is that neither reaction threatens the system that produced the image in the first place.

that's usually why context is always missing. you never really hear whether products were gifted, returned, written off, expired, unused, etc. or whether the person is a professional, sponsored, or staging excess for attention. ambiguity is the point; confusion keeps you arguing in the comments instead of questioning why your questioning in the first place.

and i understand! professionals may need an excess of items to get their jobs done, but the system doesn't target professionals. it targets everyday people like you and me.

late stage capitalism doesn't need you to want these things, but it just needs you to be emotionally activated by them. outrage is just as profitable as aspiration.

so when people say "this is just meant to make you upset," they're right. But not in the way that they think. The issue isn't that you felt something, but more so that you're trained to emotionally respond to excess instead of critically interrogating it.

ai feels invisible because you only see the final answer on your screen, not the giant physical system running underneath it. every prompt relies on servers, cooling systems, power grids, rare metals, and nonstop energy use. ai may feel like vapor, but its environmental impact is very real.

ai uses a lot of resources because modern models are so large. training them can take weeks of nonstop computation. this requires electricity, water, hardware, and maintenance. even everyday use consumes meaningful energy because data centers must stay online at all times to respond to billions of requests.

researchers usually agree that ai’s carbon footprint comes from four main areas.

training is the most energy intensive part because the models need to process enormous amounts of data.

inference is the energy used every time someone asks a question or generates content. it is smaller per use but happens millions of times per day.

data centers generate heat and need huge amounts of water and electricity to stay functional, so cooling is required

chips require mining, metals, manufacturing, and shipping.

the energy impact depends heavily on where data centers are located. regions powered by fossil fuels create higher emissions, while regions with renewable energy lower the impact. water usage matters too because many cooling systems rely on evaporative cooling, which consumes large amounts of water each day. ai is not unique in this; streaming, gaming, and cloud storage also use these systems, but ai is growing fast enough to raise sustainability concerns.

carbon markets come into the picture as a tool governments and companies use to reduce emissions. a carbon market lets organizations buy and sell allowances for the amount of carbon they emit. when ai companies participate, it can push them to reduce energy use or switch to renewable sources in order to stay under emissions limits. carbon markets also raise funding for environmental projects that theoretically offset the emissions produced by training and running models. the challenge is that offsets do not always match real-world reductions, and not all carbon credits are equally effective.

making ai more sustainable is possible. there are several strategies researchers and companies are exploring.

designing smaller and more efficient models that require less computation.

powering data centers with renewable energy instead of fossil fuels.

building data centers in cooler climates to reduce the need for artificial cooling.

improving chip design so each unit uses less power during training and inference.

using liquid cooling systems that conserve water and energy.

requiring transparency reports so users know the environmental cost of the tools they rely on.

i’ve been thinking about how much weight society, and colleges especially, put on gpa in the admissions process. and honestly, i feel like we need to chill with it.

first off: colleges are businesses. it sounds blunt, but it’s true. a college isn’t just a dreamy place for eager students to learn. it runs like an institution with goals. they want future changemakers, people who’ll be notable alumni, people who’ll help build the school’s brand, network, influence. so if you’re a college admissions office, what do you do? you recruit the folks who check the obvious boxes: high gpa, high test scores, “safe” extracurriculars. you want someone who’s done everything “right.”

but here’s the kicker: by putting so much weight on gpa, you’re basically pulling in a bunch of people whose main goal has been “get a good gpa.” they’ve learned the game. they’ve optimized for grades. there’s nothing inherently wrong with working hard to get good grades, but the problem is: their purpose might be limited to that. when they behave purely as students whose goal is “maximize my gpa,” or "get into an ivy league school" you lose the person whose goal is “find a problem, fix a problem, uproot something, make something new.” you lose the passionate person. the risk-taker. the person who maybe got a lesser gpa because they were busy creating something, experimenting, failing, learning.

so if a college says “we want changemakers,” but then questions students with a "gpa below x", there’s a mismatch. you end up rewarding “safe students” rather than “impact students.”

second point: yes, there have been shifts: more holistic admissions, more weight on essays, extracurriculars, diversity of experience. but i worry we’re sliding back to gpa‐obsession. for example: i read about brown university entering a voluntary federal agreement in 2025 that required disclosure of admissions data (including grades) and other oversight as part of restoring research funding.

the fact that federal funding deals incorporate admissions metrics signals an external pressure to quantify “merit” in ways that almost always favor high gpa. it’s upsetting because brown is known, at least in reputation, for intellectual curiosity, creative thinkers, exploration. but when external systems push “grades first,” the door closes to those who may have lower gpa but higher potential for impact.

think about the kid who spent summers volunteering, building something from scratch, maybe took on leadership in non-traditional ways, maybe got a few Bs because they were juggling passion projects and classes. that kid may have the spirit colleges say they want, but then they’re cut because of gpa. that hurts.

it sends the message that the "safe" path (study hard, hit all A’s, take a bunch of APs) is the only path. it discourages risk, it discourages exploring. “if you can’t guarantee an A, maybe don’t try that big project.” that stifles creativity.

so what do we do instead? or at least what should admissions offices do?

weight gpa less. yes, still consider it; it reflects persistence, ability to meet deadlines, handle academics. but don’t let it dominate to the point that it filters out everyone else.

give meaningful weight to curiosity, creativity, resilience, risk-taking, leadership in non-traditional spaces. show that you value the student who tries to fix something, not just the student who avoids failing.

include more narrative space in applications where students can show what they did and learned, not just what grade they got. maybe even include questions like “what project did you do, what problem did you take on, what happened when you failed?”

for the public and policy side: stop legitimizing “grades = merit” as the only metric. create accountability for admissions systems to reflect the real diversity of talent.

first, i want to address the fact that i've barely posted in the past few months. i'm sorry, my schedule was just been so packed with school, i forgot to update this blog! i'll try to be more consistent now.

anyway, girl math isn’t people being bad with money. it’s people being textbook examples of how behavioral economics works in real life.

the whole “if i pay in cash it’s free” or “i returned something for fifty so this sixty dollar thing only cost ten” logic sounds silly, but it’s actually kind of genius. it’s not a financial mistake, it’s data. it’s proof that we don’t spend based on numbers, we spend based on feelings, stories, and tiny mental shortcuts that make sense only to us.

behavioral economics was built on this idea. it says people don’t make perfectly rational choices; they make human ones. we value emotions more than math. we separate money into imaginary categories. we justify, compare, and reframe until a purchase feels right.

girl math is that theory in action. it’s mental accounting when we treat cash like it’s “not real money.” it’s loss aversion when we chase discounts just to avoid feeling like we “missed out.” it’s anchoring when a $300 coat suddenly looks affordable next to a $600 one. it’s instant gratification when future-me gets sacrificed so present-me can feel joy right now.

every behavioral economist should be thanking the internet for girl math. it’s a living case study, wrapped in humor and self-awareness. people are literally turning cognitive biases into memes. and what’s wild is that it proves the science works; the same patterns researchers wrote papers about decades ago are now trending audio on tiktok.

girl math isn’t financial literacy. it’s not trying to teach you how to budget. it’s showing, in real time, how our brains bend logic to protect our emotions. it’s the theory of behavioral economics playing out in public, in a way that’s funny, relatable, and way more honest than any econ textbook.

i tried making this poem to nails how we all kinda feel about social media and the constant feed of stories and pics. it’s like being full but still starving, you know? how the scroll never stops and somehow controls us more than we control it.

lines like “we’re swept away by the current, adrift, lost in the noise” really hit different when you think about how much time we spend glued to screens.

it’s not just about the addiction either. it’s about how those perfect, filtered lives start feeling like our own, and then what’s left of us?

it made me stop and think about how easy it is to forget what real hunger feels like when you’re always snacking on the fake stuff.

link to poem: https://www.instagram.com/p/DKscoKtzVLd/?img_index=4

nonprofits are weird sometimes—like, if they’re not making profit, how do they even keep going? do they just live off donations?

not exactly. nonprofits do make money, but they use it differently than for-profit businesses. let’s break it down.

not about profit, about purpose

the main difference between nonprofits and regular businesses is the end goal. for-profits want to maximize profit. nonprofits want to maximize impact. the “extra” money they earn doesn’t go into shareholders’ pockets—it goes right back into keeping their mission alive.

so instead of focusing on how much money they make, nonprofits focus on how much good they do per dollar spent.

where the money actually comes from

nonprofits don’t rely on selling products the same way corporations do. instead, they fund themselves through...

grants (from government or private institutions)

donations (from individuals, communities, or fundraising drives)

events + campaigns (galas, auctions, walks, etc.)

services or products (some nonprofits do sell things, like merch, museum tickets, or classes, but the money goes straight back into the cause)

this mix of income keeps them afloat even without traditional profit margins.

the expenses are real though

nonprofits still have to cover basics: paying employees, covering rent, tech, insurance, retirement funds, and all the boring behind-the-scenes stuff. the difference is they don’t aim to exceed those costs for personal gain. breaking even while continuing to fund their programs = success.

the bigger picture

nonprofits survive because their value isn’t measured in profit but in impact. keeping the lights on is important, but their true “survival” depends on showing results: feeding families, educating kids, preserving art, saving animals. that’s how they secure ongoing support and funding.

nonprofits don’t survive without profit—they survive without personal profit. every dollar is recycled back into the mission. and while the money structure looks different, the hustle is just as real. SOURCES

Segal, Troy. "How Do NGOs Get Funding?" Investopedia, 30 Jan. 2023, www.investopedia.com/ask/answers/13/ngos-get-funding.asp. Accessed 31 Aug. 2025.

"7 Ways a Nonprofit Can Generate Revenue." 501c3-Nonprofit.org, 3 Jan. 2022, www.501c3-nonprofit.org/blog/7-ways-a-nonprofit-can-generate-revenue. Accessed 31 Aug. 2025.

"How Do Nonprofits Receive Grants?" Instrumentl, www.instrumentl.com/blog/how-do-nonprofits-receive-grants. Accessed 31 Aug. 2025.

"Chapter 11: Revenue Streams for Nonprofits." Jotform, www.jotform.com/help/485-chapter-11-revenue-streams-for-nonprofits. Accessed 31 Aug. 2025.

"Nonprofit Revenue Diversification: What It Is and Why It Matters." Jitasa, www.jitasagroup.com/jitasa_nonprofit_blog/nonprofit-revenue-diversification. Accessed 31 Aug. 2025.

Santana, Aamer Madhani, and Mariam Fam. "US Charities Are on the Brink as Government Funding Dries Up and Needs Soar." AP News, 30 Apr. 2024, apnews.com/article/ed2e5b30445c9ffdb07346e42c0abfa3. Accessed 31 Aug. 2025.

"Why You Need More Than One Source of Revenue for Your Nonprofit." Get Fully Funded, www.getfullyfunded.com/more-than-one-source-of-revenue-for-your-nonprofit. Accessed 31 Aug. 2025.

"10 Revenue Streams for Nonprofits." Wild Apricot, www.wildapricot.com/blog/nonprofit-revenue-streams. Accessed 31 Aug. 2025.

in basic econ, demand comes from consumers--we decide what we want, and businesses react by supplying it. simple, right?

but in real life, things are a little messier. like, did we all just randomly decide we needed oat milk in 2020? or did brands flood our feeds with creamy, aesthetic coffee content until we caved and made it our personality?

how brands shape demand (on purpose)

brands don’t just sit around waiting to see what people want--they create desire. they do it through...

advertising (those painfully specific tiktoks that hit too close to home)

trend forecasting (predicting what vibes will be in 6 months and prepping products early)

influencer culture (if your favorite influencer wears it, you’re 67% more likely to consider it)

scarcity tactics (limited drops, collabs, “only 3 left” warnings = instant fomo)

they introduce something new, make it cool, and suddenly we all feel like we need it. that's called manufactured demand--the product didn’t solve a need, it created one.

on the flip side, brands also watch us like hawks. our google searches, shopping habits, and tiktok comments literally shape what they do next. we are walking data mines.

when enough people start wanting something (like more sustainable packaging or gender-neutral sizing), brands pivot to meet that demand. in this case, we’re in charge. the demand already existed--they just didn’t want to miss out on the $$$.

this is reactive branding, and it happens when ignoring the consumer would be a bigger risk than changing direction.

truth is, demand isn't a one-way street. brands create demand, but they also follow it. it’s a cycle.

consumers want something

brands notice it or amplify it

they market the hell out of it

more people want it

repeat

sometimes brands lead. sometimes they chase. sometimes it’s impossible to tell the difference.

who's really in control? the answer kind of depends on how media-literate we are. if we don’t pause and think about why we want the things we want, brands will absolutely run the show.

but if we’re intentional--about what we support, buy, or hype up--then we shift power back into the hands of the consumer. the demand becomes real, not manufactured. and brands? they’ll adjust, because they have no choice.

brands are both creators and followers of demand. they feed off trends but also manufacture them. and we, the consumers, are both influenced and influential.

when we talk about how businesses function, most fall into two main categories: capital-intensive or labor-intensive. it might sound like textbook jargon, but this divide lowkey shapes the entire economy. from the brands you buy to the jobs you’ll have one day, knowing the difference helps you see who’s really running the world behind the scenes.

what is capital-intensive?

capital-intensive businesses are the ones that rely heavily on money, machines, and technology. they need big upfront investments in physical stuff like factories, vehicles, high-tech equipment, or infrastructure to even exist. we’re talking oil companies with massive rigs, tech giants with data centers, or manufacturing plants with fully automated production lines.

these businesses spend a lot early on, hoping to make it back later. this is called speculative cost. they’re betting on future success by dropping millions (or billions) into machines and systems now. think companies like exxonmobil, boeing, or tesla.

what is labor-intensive?

labor-intensive businesses, on the other hand, depend more on people. their biggest expense isn’t fancy tools or equipment. It’s the workforce. they need skilled, motivated humans to do the work. restaurants, salons, tutoring services, construction crews, even indie clothing brands fall into this category. they’re built around time, care, and actual hands-on effort.

these businesses might not have the flashiest tech, but they’re adaptable. they can scale up or down quickly, especially in response to changing demand or new trends. plus, they’re often the most connected to community and culture.

who runs the world?

both. but in very different ways.

capital-intensive industries power the backbone of modern life (energy, transportation, infrastructure, global manufacturing). they’re the big players, the heavy lifters. without them, a lot of what we take for granted (like having a phone, car, or lights on in your house) wouldn’t exist.

but labor-intensive industries shape the day-to-day. they’re the ones you interact with constantly. they create the vibe. they’re also responsible for most of the service jobs, the creative hustle, and the human experiences we actually remember. both are essential. they just run the world in different lanes.

why it matters

knowing the difference isn’t just academic; it impacts everything. it affects how businesses respond to inflation, how job markets shift, and even how sustainable or risky a business model is. capital-intensive businesses can be harder to start but have huge scalability. labor-intensive ones are more personal and flexible but depend heavily on people’s time and energy.

with automation and ai creeping into more spaces, the lines between these two are starting to blur. a restaurant might use robots to flip burgers. a tech company might rely more on creators and social media managers. the balance is always shifting.

the bottom line

it’s not about one being better than the other. it’s about understanding how they work together. the economy needs machines and people, money and time, tech and touch. and whether you’re planning a career, starting a business, or just want to sound smart in econ class, knowing how the world runs behind the curtain is a pretty solid flex.

SOURCES:

The Investopedia Team. Investopedia, 20 Feb. 2024, https://www.investopedia.com/terms/c/capitalintensive.asp.

Kenton, Will. Investopedia, 28 Nov. 2020, https://www.investopedia.com/terms/l/laborintensive.asp.

everyone’s selling identical products. no one controls prices. there are zero barriers to entry. everyone knows everything (perfect info), and profits are just enough to survive; no one’s balling out. it’s the most “fair” version of capitalism that exists on paper.

but here’s the catch: it only exists on paper.

in real life, markets are way messier.

perfect competition assumes conditions that rarely ever happen:

products are identical (but let’s be real; even water comes in 47 brands and 12 aesthetics)

there are tons of buyers and sellers (try entering the airline industry lol)

everyone has the same info (but marketing literally exists to prevent that)

firms are price takers (hi again, Apple, charging $1,200 for a phone because they can)

and those conditions break down fast. take the cereal aisle. technically, all you need is oats and sugar, but brand loyalty, ads, packaging, and flavor inflation turn it into a profit warzone. welcome to monopolistic competition, where companies sell similar but differentiated products and compete on vibes, not just price.

or take energy markets. just a few massive players, high startup costs, and everyone still needs the product. that’s not perfect competition. that’s oligopoly territory.

so yeah, perfect competition is a helpful model. but it’s just that. a model. like those pinterest study routines. nice in theory, doesn’t always survive contact with reality.

and that’s okay. just means we have to think critically. ask better questions. look past the graphs.

so let's say it's friday night. you have $20, two invites, and the illusion of free will. option one: go out for ramen with my friends. option two: finally buy that book you've been eyeing (and then dramatically annotate it). you can’t do both unless you magically generate more money (which, sadly, the federal reserve won’t let you do).

enter opportunity cost. aka, the thing economics uses to remind you that every choice comes at a cost. not just in money, but in whatever you had to give up.

if you choose ramen: you lose your cozy solo night and the book. if you choose the book: you lose the serotonin of slurping noodles with people you love. either way, you lose something.

but it gets worse.

opportunity cost doesn’t just ruin your friday night — it’s everywhere. sleep in or study? splurge on iced coffee or save for gas? take ap stats or “enjoy your life” (according to my friends)? these aren’t just decisions. they’re little economic experiments. and we’re all participating, whether we like it or not.

and here’s the kicker. not choosing is still a choice. if you sit on your bed for two hours overthinking both plans, then guess what? you've already paid the opportunity cost of not doing either. rip.

so what’s the point?

well. understanding opportunity cost doesn’t mean you’ll always make the “perfect” decision. but it helps you make intentional ones. it reminds you to think:

“what am i giving up by saying yes to this?” “is it worth it?” “am i making this decision, or is my brain on autopilot again?”

business isn’t just branding and buzzwords; it’s economics in action. behind every strategic decision, there’s a blend of numbers, human behavior, and market dynamics. here’s how real businesses use key economic principles to guide their moves.

before launching a new product, opening a store, or hiring another employee, businesses ask: is the benefit worth the cost?

a cost-benefit analysis weighs the expected gains (like revenue, efficiency, or customer satisfaction) against the potential costs (money, time, risk). it’s how businesses avoid waste and prioritize what actually moves the needle.

example: should a bakery open on sundays?

costs: employee wages, utility bills, potential burnout

benefits: extra sales, more brand exposure, convenience for customers

if the benefits outweigh the costs long-term, they go for it. if not, they don’t. it’s that simple (and that powerful).

price isn't random. it’s a reflection of how much people want something (demand) and how much of it exists (supply). businesses constantly study these patterns to make smart decisions.

example: if demand for electric cars is skyrocketing and supply is low, prices go up. that signals automakers to ramp up production or innovate faster. if the market’s flooded with a product no one wants? prices drop, and companies pivot.

this also helps with setting prices that customers will actually pay, avoiding overproduction and waste, and knowing when to restock, scale back, or promote more

every business has limits: time, budget, people. so they have to choose. and every choice means not doing something else.

that’s where opportunity cost comes in; the value of the next-best option you’re giving up.

example: should a startup spend $10k on marketing or new tech? if they choose marketing, the opportunity cost is the growth they could have seen with better tech — and vice versa.

smart businesses calculate these trade-offs constantly. they don’t just ask “can we do this?” they ask, “is this the best use of our limited resources?"

marginal analysis: small changes, big insight sometimes the most strategic decisions come from small tweaks. marginal analysis means asking: what’s the added cost and benefit of just one more?

should we hire one more employee?

produce one more unit?

spend one more hour training staff?

example: if adding one more server during lunch rush boosts revenue more than it costs in wages, it’s worth it. if not, skip it.

businesses that understand economics don’t just guess, they adapt. they know how to read the room (and the market), choose intentionally, and build sustainable strategies.

so next time you see a product launch, a store closing, or a price increase, know that behind it, there’s likely an economic principle at work.

SOURCES: "Cost-Benefit Analysis." Investopedia, 14 July 2021, https://www.investopedia.com/terms/c/cost-benefitanalysis.asp.

"Law of Supply and Demand." Investopedia, 28 Sept. 2020, https://www.investopedia.com/terms/l/law-of-supply-demand.asp.

"Opportunity Cost." Investopedia, 4 Oct. 2021, https://www.investopedia.com/terms/o/opportunitycost.asp.