Writing & Research

Technology, ethics,
& the human impact

Exploring how AI systems shape communities and what it means to build safe and trustworthy technology

 Scroll to explore
spin
click to learn about me

I spent most of my time writing in college

I would be remiss to have them never see the light of day — so here are some of my favorites.

Honors Thesis · Philosophy & Data Science
World Models
NYU · Advisor: Ned Block

LLMs struggle with complex reasoning and conveying uncertainty. This thesis examines world models as a solution — comparing LeCun's empiricist JEPA with Tenenbaum's nativist probabilistic language of thought.

Download PDF
Philosophy · Ethics of AI
Does Surveillance Capitalism Diminish Autonomy? A Case Study Using Cambridge Analytica
How surveillance capitalism diminishes autonomy through unknowing manipulation

Cambridge Analytica harvested 87 million users' data to build 253 predictive algorithms, targeting 'persuadables' with fear-based ads. A case study in how behavioral surplus turns citizens into pawns.

Read essay
Philosophy · Ethics of AI
Do AI Systems Wrongfully Discriminate?
On algorithmic hiring, demeaning treatment, and structural subordination

Drawing on Hellman, Moreau, and Anderson to evaluate why certain forms of algorithmic sorting in hiring are not just discriminatory — but unjust.

Read essay
Philosophy · Existentialism
Can One Imagine Beth Harmon Happy? A Camusian Reading of the Queen's Gambit
On the Queen's Gambit, the absurd, and why the last scene changes everything

Beth Harmon spent her life using chess as a means to an end — until the final scene, where she plays for no reason at all. A reading of the Queen's Gambit through Camus's revolt against meaninglessness.

Read essay
Philosophy of Science · AI
Are Artificial Neural Networks Revolutionary Under Kuhn's Framework?
Are artificial neural networks actually revolutionary? Kuhn would say not so fast.

Everyone calls neural networks revolutionary. But applying Kuhn's framework to cognitive science reveals that connectionism hasn't replaced classical computationalism — the paradigm shift is still pending.

Read essay
Philosophy of Science · Data Science
Does Popper's Falsification Survive in Modern Data Science?
How hypothesis testing diverges from Popper's falsification

Data science falsifies the null to corroborate the alternative. Popper would never. An investigation into whether modern hypothesis testing survives Popper's methodology.

Read essay
Philosophy · Political Philosophy
Does the General Will Make Us Free? Rousseau on Sovereignty and Moral Freedom
Rousseau's argument for sovereignty and the circularity problem underneath

Rousseau says you need the general will to be free, but you need freedom to form the general will. A reconstruction of his argument — and the dilemma at its foundation.

Read essay
Philosophy · Political Philosophy
Are Wage Workers Free? Cohen on Collective Unfreedom and the Limits of Capitalism
Cohen's key analogy, collective unfreedom, and why Anita can't grab the key

Cohen says wage workers are individually free but collectively unfree. His key analogy is elegant — but it rests on the fickle assumption that most people in the room don't want to leave.

Read essay
Philosophy · Feminist Theory
Who Gets to Be a Woman? Haslanger and Jenkins on the Ameliorative Concept of Gender
Haslanger's subordination account, Jenkins's internal map, and trans exclusion

If gender is defined by subordination, trans women who aren't perceived as women are excluded from the category entirely. Jenkins proposes that self-identification and class must work together.

Read essay
Philosophy · Metaphysics
Is Race a Social Kind? Haslanger's Constructionist Account
Sally Haslanger's social constructionist account of race

Race works like a recipe: pick a phenotypic trait, assign it to a group, attach negative characteristics, subordinate. The folk concept is more useful for justice than any genetic cluster.

Read essay
Philosophy · Ethics
Is Aristotle's Ethical Theory a Plausible Ethical Theory?
Why Aristotle's virtuous agent can never truly act for another's sake alone

Can an ethical theory demand sacrifice without motivation? Aristotle says the excellent person will die for their friends — but only because it achieves the fine. Every act circles back to the self.

Read essay
Philosophy · Argument Reconstruction
A Logical Reconstruction of What Justice Is
Premises, conclusions, and Socrates vs. Thrasymachus on justice

Fourteen premises, one conclusion: rulers cannot seek their own advantage. A formal reconstruction of the Republic, Book 1 — turning Plato's dialogue into deductive logic.

Read essay
Commentary · Law & Politics
Dark Money's Hand in the Supreme Court
How billion-dollar donations to conservative think tanks stacked the Court

A $1.6 billion donation to the Federalist Society. Justices groomed since law school. An originalist majority that overturned Roe, gutted the EPA, and replaced longstanding legal tests — all traceable to the same dark money pipeline.

Read essay

Academic Publications

Research conducted at NYU's Center for Data Science and Information Law Institute, under the supervision of Umang Bhatt and Katherine Strandburg.

Research Paper
Unequal Uncertainty: Rethinking Algorithmic Interventions for Mitigating Discrimination from AI
Holli Sargeant, Mackenzie Jorgensen, Arina Shah, Adrian Weller, Umang Bhatt

A legally informed analysis examining risks of discrimination associated with selective abstention in AI-assisted decision-making. Proposes a framework integrating AI assistance with selective frictions to mitigate risks and enhance fairness.

Conference Paper · AIES 2025
Documenting Deployment with FABRIC: A Repository of Real-World AI Governance
Mackenzie Jorgensen, Kendall Brogle, Katherine Collins, … Arina Shah, … Umang Bhatt

Contributed to a public repository of real-world AI deployments and oversight patterns, based on interviews with practitioners.

View publication
Work in Progress
Prompting Large Language Models to Verbalize their Uncertainty with RAG
NYU Center for Data Science

Investigating how Retrieval Augmented Generation can modulate the expression of uncertainty in LLM outputs.

View PDF

Books That Shaped My Thinking

I have so many more favorites (Fahrenheit 451, The Book Thief, Tomorrow, and Tomorrow, and Tomorrow, Anna Karenina) but if I had to pick 10, these are the ones I would wholeheartedly recommend — and the ones that brought me out of some pretty bad reading slumps.

Read why I love these
01
The Namesake
Jhumpa Lahiri
The first book that put the emotions behind cultural clashes into words.
The experience of reading this book felt like slicing through butter. The words, the plot, the movement through Gogol's life felt seamless yet impactful. I'm the daughter of immigrants, but it wasn't until high school after joining the golf team that I felt significant cultural clashes. I had what my family notoriously called my "brown friends" and "white friends," where I apparently code-switched more than I knew. It wasn't anything negative — I love my friends and they are still the people closest to me today — but I did feel like I was swapping identities at times. The Namesake was the first book that put the emotions behind cultural clashes into words.
02
Educated
Tara Westover
A reminder of what a privilege it is to have an education.
I read this right after I graduated college and was grappling with the importance of education. I felt a bit defeated that this shiny degree I had earned had so little value in a job force that was choosing LLMs and agents over real graduates. With the echo chamber of Instagram also telling me that education was pointless, I began to believe it — until I read this book and realized what a privilege it is to have an education.
03
Empire of AI
Karen Hao
Who gets to govern AI? This book asks the question that won't leave me alone.
I just want to start off by saying that Karen Hao is incredible. This book is extremely well researched and very clearly outlines the stakes of the rapid development of AI. This book is what got me to start posting about tech-related content on Instagram in order to hopefully stir up some conversations about what rapid development means and make the concepts and consequences more accessible to people outside of the tech world. This book asks an important question: who gets to govern AI? It made me think deeply about the race to "AGI" — Sam Altman said that AGI will save humanity, but in trying to achieve it, will it destroy it too?
04
The Kite Runner
Khaled Hosseini
The first book that made me realize what reading really does — it lets you peek into a world that isn't yours and empathize with it.
This is a classic. I admittedly read this way too young, in 7th grade, and was extremely traumatized. Then I re-read it in high school and then in college. Each time I read it I feel like I grew up a little — I thought I had stopped mentally maturing at 15, but when I re-read this book in college I definitely realized I had. The themes of friendship and betrayal hit differently each time. I always viewed Amir as evil, but when I read it in college I felt sympathy; it's not easy doing the right thing, especially when you're scared. This was one of the first books that made me realize what reading really does — it lets you peek into a world that isn't yours and empathize with it.
05
Empire of Pain
Patrick Radden Keefe
A psychological thriller that happens to be real.
This was the first investigative journalism book I read and I could not stop talking about it for years — I think I was insufferable to my friends. The Sacklers, and not only the empire they built but the lies they produced in order to build a billion-dollar empire, astonished me. From building a fake pain association, to doing inside deals with the FDA, to causing a full-blown opioid crisis — it will always be shocking. The way this story is narrated and written feels like a psychological thriller that happens to be real. It was also really terrifying to see someone who was a physician, who took the oath of helping people, turn into a monster in the pursuit of billions.
06
The Bell Jar
Sylvia Plath
If you are a twenty-something girl, I would highly recommend the read.
I read this when I was a junior in college. I had just begun my glamorous New York lifestyle. Unlike freshman and sophomore year when I was a quiet nerd, I decided I wanted to experience the Gossip Girl version of New York. I started having glamorous nights out, but definitely not without a cost. I was also trying to figure out what I wanted to do with my life — go to law school? Finance? Consulting? PhD in philosophy? I felt connected to Sylvia because of the societal pressure she faced. Although different from mine, I understood the weight of expectations: everyone saying you have potential but the possibility of never living up to it. If you are a twenty-something girl, I would highly recommend the read.
07
A Psalm for the Wild-Built
Becky Chambers
My comfort read. You can finish it before your coffee gets cold.
I gave this to my best friends in college when we graduated. It's about a world post-AI where intelligent robots are sent off into the wilderness and separated from the human population. A monk from the town ventures out and becomes friends with one of the robots. It's an optimistic science fiction book (I know, never heard of) that you can honestly finish before your coffee gets cold. Every time I read it I feel content and light, like things aren't that serious, you can still have whimsy, and new relationships are worth exploring. It's definitely my comfort read.
08
And Then There Were None
Agatha Christie
The best kind of read is the kind that keeps you up until 3 a.m.
This is a classic — one of the first real mysteries I read after Geronimo Stilton. Mystery is definitely my favorite genre and I owe it all to Agatha Christie. If you are in a reading slump I highly recommend reading anything written by her; it will keep you up until 3 a.m., which I think is the best kind of read. I honestly really like how interactive it is, like trying to figure out who did it but always saying, well no, it can't be that obvious. I always got it wrong and was usually shocked by the ending. I also really love the ensemble of characters and how they interact with each other — you get so many different backstories and I find that so intriguing.
09
The Age of Surveillance Capitalism
Shoshana Zuboff
The book that sparked my interest in the negatives that can arise from technology.
This was the book that sparked my interest in some of the negatives that can arise from technology. I read this my sophomore year of college when ChatGPT launched, and it inspired several of my philosophy papers on autonomy, privacy, and data as a resource with comparative value to that of water. The Pokémon Go case study has also stuck with me for years.
10
The Rose Code
Kate Quinn
The countless women who had to keep their stories about working in British intelligence a secret.
I am a huge fan of Kate Quinn and have read many of her other books that focus on women in World War I or World War II. Another one of my favorites is The Briar Club. This one focuses on three different women who worked at Bletchley Park: one was a German translator, another operated the machines to decrypt Enigma codes, and another was a codebreaker. I learned so much about the roles that women played during World War II, about the probably countless women who had to keep their stories about working in British intelligence a secret, and the trauma that followed them from this work.

Arina Shah

In pursuing two, in my opinion, polar opposite degrees — Philosophy and Data Science — while simultaneously building products and reporting on recent developments in AI, I've come to realize these fields are far less disparate than they appear, and that the most pressing questions in AI demand interdisciplinary effort. I'm interested in how we can innovate responsibly, building trustworthy AI systems that emphasize human-machine collaboration rather than replacement.

My research sits at the intersection of AI and the law. I started by examining GDPR compliance and explainability mechanisms — what ought to be legally required when AI systems make decisions about people, from Medicare determinations to hiring. More recently, I co-authored two papers: one assessing interventions for making AI systems more equitable and proposing a new framework, and another documenting the governance mechanisms that AI enterprises actually use in practice. My honors thesis in philosophy, advised by Ned Block, examined world models as an alternative architecture for AI reasoning.

That same interest in collaboration led me to co-found Fruition. After building a research marketplace to help universities streamline finding relevant research opportunities and make it easier for faculty to sift through applications, I noticed a broader pattern: students wanted to innovate and explore, but not in isolation. That's why we built Fruition — a platform that connects young founders with the right teammates. It also reflects my belief in human-machine collaboration. Matches aren't solely curated by algorithms; we use our own touch working alongside the system to produce the best pairings for our innovators, exploring the psychology of what makes people work well together and allowing them to connect quickly.

Outside of my academic pursuits, I volunteer with Northwest CASA for sexual assault prevention, play golf, box, and dance Kathak. All things considered, I'm a 22-year-old trying to make sense of a world that's changing faster than most of us can keep up with, and doing my best to make sure the people building what comes next are building it thoughtfully. If you're interested in joining Fruition or chatting about any of this, I'm always open to doing that over an oat milk latte 🎀

Education

New York University
B.A. Philosophy & Data Science

Research

NYU Center for Data Science
Information Law Institute

Focus Areas

AI & Technology Law · Algorithmic Fairness
AI Governance · Digital Safety

Connect

LinkedIn · Instagram · Email

For Builders
Young, ambitious, and looking for your co-founder?

Fruition is an AI-powered platform I co-founded to connect early-stage entrepreneurs with the right team members.

Beth Harmon, the protagonist of "Queen's Gambit," is a fictional chess prodigy who rises to fame by winning the Soviet Chess Championship against the world's leading player, Vasily Borgov. Beth's early life is marked by trauma; she becomes an orphan at nine following her mother's fatal car crash. At the Methuen orphanage, she discovers chess through the janitor, Mr. Shaibel, and becomes addicted to both the game and the tranquilizers given to the children. Beth is then adopted by the Wheatleys. While attending public school, she finds herself a pariah, with her drab clothing and unconventional personality. Beth begins competing in tournaments. Her chess prowess earns her both money and recognition, fueling a burgeoning career that leads her across the country. Despite external success, Beth battles inner demons, continuing to indulge in tranquilizers, alcohol, and smoking. Beth's challenges peak during an international tournament where, after her adoptive mother's sudden death and losing a game against Borgov, she spirals deeper into addiction. In the series finale, Beth decides to travel to Russia for a rematch against Borgov. Victorious, Beth chooses not to pursue the celebrity lifestyle expected of her but instead to play chess in the park with elderly strangers. I will show that Beth Harmon lives a good life according to Camus by assuming that in the last scene, she realizes the absurd and embraces it, thus revolting against the meaninglessness of life.

Albert Camus believes that a good life is one that embraces the absurd, "Living is keeping the absurd alive (...). One of the only coherent philosophical positions is thus revolt. It is a constant confrontation between man and his own obscurity (Camus,18)." The absurd can be thought of as a paradox. A person's deepest desire is to understand the world and find meaning; however, grasping meaning is beyond human abilities making the world, as a person would interpret it, meaningless. Camus differentiates himself from other existentialists like De Beauvoir, Sartre, and Kierkegaard, because he believes the absurd is tied to the human condition and therefore inescapable, "A man who has become conscious of the absurd is forever bound to it (Camus, 11)." Other existentialists believe that the world is meaningless as well, however, there is hope beyond the world. Believing in God, the One, or creating your meaning can allow a person to overcome meaninglessness. Camus believes that these are all delusions.

The main question and the only question that Camus views of philosophical importance is that of suicide. Camus believes that a person should not commit suicide, "Suicide is repudiation (Camus, 19)". Suicide would mean succumbing to a meaningless life. By committing suicide, one would be escaping the absurd rather than accepting it. However, what makes a life good is embracing the absurd. It is not enough to embrace the absurd just once. A good life for Camus would be one where a person comes to the realization of the absurd and constantly embraces it by continuing to live as a revolt against meaninglessness.

Camus illustrates what a good life looks with Sisyphus. Sisyphus is a god who is condemned to pushing a boulder up a hill for the rest of his life. However, Camus argues that Sisyphus is living a good life because he has realized the absurd and practices it by revolting against the condition of existence itself by continuing to live. Once Sisyphus embraced the absurd the most fulfilling way to live was in continuous engagement with the present moment. Rather than seeking instrumental gain or dwelling on the past, Sisyphus enjoys each moment of pushing the boulder. The painful and traumatic punishment that Sisyphus faces is what allows him to recognize the absurd and revolt. Camus goes on to say that there is a fundamental link between experiencing the absurd and achieving happiness: "Happiness and the absurd are two sons of the same earth. They are inseparable (Camus, 24)." By accepting the absurd Sisyphus is liberated from the despair that arises from seeking meaning, this freedom allows him to be happy. Therefore Camus concludes that "one must imagine Sisyphus happy (Camus, 24)."

Beth similar to Sisyphus is dealt bad cards in life. Losing her mother twice, and being alone and dejected by those around her seems like a punishment. When Beth loses her biological mother at the age of 9 she realizes that the world is meaningless. We can see this from her loss of interest in studying, speaking to other people, and making friends. Further, Beth is not shown as being sad but rather empty. However, upon first seeing the chess board, she seeks to create her own meaning through the game. At this point, Beth has realized the absurd but rather than embracing it, she deceives herself by thinking that she can create her own meaning. During her formative years, chess has instrumental gain for her. She uses it to escape the trauma of her mother dying, to escape being alone, for fame and recognition, and to give her purpose. Beth could be seen as different from Sisyphus because she reaches her goal of being a world champion and beating Borgov, while Sisyphus never really reaches the top of the mountain. However, this goal is not the top of the mountain for Beth. At the end, Beth is sitting in the back of a taxi with her bodyguard who is telling her of all the extravagant events awaiting her back in the U.S.; meeting with the president, an interview on the Tonight Show, all her friends, etc. Beth hops out of the cab, missing her flight. The show concludes with her smiling while playing chess with the elderly in a park.

The very pointlessness of this act, playing chess knowing she is not going to gain anything from the act, is the moment Beth embraces the absurd. She comes to value chess for the sake of itself and not as a means to an end. I would argue that Beth does this because after achieving her goal she realizes that there is no way to escape the meaninglessness of life. Rather than turning to something else to give her purpose, a new craft, God, or indulging in destructive acts like drinking, smoking, doing drugs, or committing suicide, Beth revolts by continuing to live doing the thing she has always done: playing chess, however this time without looking for a further gain. This makes her, like Sisyphus, free, as she is not trying to find meaning through chess anymore.

I agree with the theory that embracing the absurd and revolting by continuing to live is a good life. The anxious worry about the future or harboring the pain of the past causes suffering to a person. Revolting for Camus looks like living in the present and valuing every action for its own sake rather than focusing on the consequences of the actions. One may refute that the motivation to continue living comes from the consequences of our actions, the fruits of our labor. For instance, Beth studies chess rigorously in order to win the world championships. However, that adds pressure to act. If Beth were to play chess with a focus on just winning then she would chase a desire that is insatiable. There will always be new people who could beat her. Beth should study chess because she values the game for the sake of itself. While it may seem pointless to play chess for no reason, like pushing the boulder up the hill, this is a revolt against meaninglessness.

When Beth does not win her first match against Borgov she becomes extremely disheartened, spiraling into drugs and drinking. Had Beth realized and embraced the absurd at that point, she would not have spiraled. She would have just continued living life and preparing for the next tournament. However, embracing the absurd makes happiness a constant state rather than a fleeting feeling. If a person has no stake in the consequence of their action then the result of that action cannot take away from their happiness. I find this theory a refreshing way of saying that the best way to live a good life is to just live in the present. Worrying about the future or clinging to the past will only cause you suffering and pain. If one accepts that the world is meaningless and still continues to live, then life itself is valuable for its own sake. If we live like this then even the stress of trying to achieve a good life goes away. If a good life is one that does not have suffering, then a life in which nothing matters and yet you continue to live will lead to no mental suffering. Hence embracing the absurd and constantly revolting against meaninglessness by continuing to live is the good life. Therefore, one must imagine Beth happy.

In Plato's Republic book 1, Socrates and Thrasymachus enter a heated debate regarding what justice is. Thrasymachus proposes "I say that justice is nothing other than the advantage of the stronger." This leads Socrates to the question of whether something that is stronger can even be advantaged. Motivating the further question of whether rulers, who are considered stronger, can seek their own advantage. After a lengthy discussion, Socrates comes to the conclusion that rulers cannot seek their own advantage but only the advantage of their subjects. The argument goes as follows: (1) Ruling is considered a craft. (2) Rulers are considered rulers if and only if they are practicing the craft of ruling. (3) Crafts are complete and perfect. (4) If crafts are complete and perfect then they do not have deficiencies. (5) Crafts seek to advantage the thing they are applied to. (6) Crafts advantage things that are deficient in the subject of the craft. (7) Only something that is deficient seeks an advantage. (8) Something that has an advantage does not seek its own advantage. (9) Crafts have the advantage of the craft itself. (10) Crafts cannot seek their own advantage. (11) Rulers seek the advantage of their subjects. (12) Rulers have the advantage of the craft of ruling. (13) Rulers cannot seek the advantage of the craft ruling. (14) Rulers cannot seek their own advantage but only the advantage of their subjects.

Premise 1 is an inferred premise in which rulers practice the craft of ruling. Premise 2 holds that rulers must be considered rulers in the precise sense—a ruler, insofar as he is a ruler, never makes errors. A ruler is considered to be a ruler when he is practicing the craft of ruling. Premise 3 initiates the discussion of crafts and their nature—crafts are without fault or impurity. Premise 5 is defended: "And aren't respective crafts by nature set over them to seek and provide what is to their advantage." Premise 6: "because our bodies are deficient rather than self-sufficient the craft of medicine has been discovered." Crafts therefore can only advantage what is deficient. Premise 9: if a craft seeks to advantage something, and crafts are not deficient, and only deficient things seek an advantage then crafts must have the advantage of the craft itself. Premise 10: If crafts have their own advantages and something that has an advantage does not seek its own advantage then crafts cannot seek their own advantage. Premise 11: "it is to his subject and what is advantageous and proper to it that he looks." Premise 12 follows from 2 and 9. Premise 13 follows from 10 and 12. Premise 14 finally follows from 11 and 13.

However, I believe that the strength of premise 12, rulers have the advantage of the craft of ruling, seems weak. I believe that there is a difference between a ruler and the craft of ruling. A ruler seems to be the vehicle that enforces the craft of ruling. While driving can be considered complete and perfect, the vehicle that controls driving could be faulty. Through this analogy, I believe that while the craft of ruling can be considered perfect and complete, I do not believe that the vehicle enforcing the craft of ruling, rulers, can be considered perfect and complete. If rulers are deficient, then they could seek an advantage. However, Plato could argue that rulers are only considered rulers in the truest sense when they are practicing the craft of ruling.

The conclusion that rulers cannot seek their own advantage but only the advantage of their subjects serves to negate the broader claim Thrasymachus makes: "I say that justice is nothing other than the advantage of the stronger." Socrates agrees that justice is an advantage but most probably not the advantage of the stronger. Socrates argues that something can only advantage what is weaker. Further, something that gives the advantage must be stronger than its subject. A ruler then seeks to advantage what is weaker, the subjects. If the ruler is stronger and the ruler cannot seek their own advantage, then the stronger cannot seek their own advantage. This conclusion then disproves Thrasymachus's claim that justice is the advantage of the stronger.

Hiring discrimination has long been recognized as a moral and legal wrong. In high-profile cases like Coca-Cola Co. in 2001 and Gonzales v. Abercrombie & Fitch in 2003, companies were found liable for systematically excluding or disadvantaging racial minorities and women. In these cases, human decision-makers acted directly on racial or gender prejudice. But what happens when such decisions are mediated by automated systems? Since 2023, 99% of Fortune 500 companies—including Facebook, Deloitte, and Amazon—have adopted AI tools for hiring. These systems, used to screen, rank, and interview candidates, promise efficiency and neutrality. Yet scholars are wary that they may replicate or intensify existing patterns of exclusion, especially against already marginalized groups.

Is it unjust when AI systems disproportionately select men over women, and especially white men over Black women, even when both groups meet formal qualifications or when Black women are underrepresented among the "most qualified"? This paper argues that such disparities constitute wrongful discrimination when they demean individuals, subordinate marginalized groups, and violate relational and substantive equality. Drawing on Deborah Hellman's theory of demeaning discrimination, Sophia Moreau's concept of structural subordination, and Elizabeth Anderson's theory of democratic equality, I develop a framework to evaluate why certain forms of algorithmic sorting are not just discriminatory but also unjust. I begin by outlining these three theories. I then argue that both demeaning and subordinating forms of wrongful discrimination are unjust because they violate substantive equality of opportunity, which Anderson presents as essential to a just society under democratic equality. Finally, I apply this framework to two cases: Amazon's resume-screening AI, which directly downgraded resumes with references to women, and a hypothetical case in which an AI system disproportionately favors white men over Black women through proxy variables and uncertain predictions.

Both Hellman and Moreau argue that discrimination is unjust when it is not morally permissible. In When Is Discrimination Wrong?, Hellman contends that discrimination is wrongful when it demeans the individual, treating someone as if they lack equal moral worth. Equal moral worth, for Hellman, means being entitled to equal treatment as a full member of society. Demeaning is not simply harmful; it is an expressive act that signals someone is less worthy of concern or respect. Hellman argues that two conditions must be met for an act to be demeaning: (1) it must express a denial of equal moral worth, and (2) it must come from an actor, such as an institution or agent, with sufficient power or status to make that expression socially meaningful. Whether an action is demeaning depends on the social meaning of the classification used, which in turn is shaped by the history and current status of the group in question. Traits like race or gender, given their histories of oppression, often carry such meaning. Thus, when a policy signals unequal standing through these traits, it constitutes wrongful discrimination. While Hellman distinguishes demeaning from subordination, she acknowledges that the former often leads to the latter.

Moreau builds on Hellman's theory by arguing that discrimination is also wrongful when it contributes to unjust subordination—a form of group-based disadvantage that Hellman's focus on expressive disrespect cannot fully explain. While Hellman sees wrongful discrimination as a denial of equal moral worth by a powerful actor, Moreau emphasizes that discrimination can also occur indirectly through policies and practices that privilege dominant groups. In Faces of Inequality, she critiques Hellman by presenting cases, such as a white employee spitting on a Black supervisor, that show how deep social hierarchies persist even without institutional authority. Moreau argues that we must understand discrimination through group-based social positioning, where certain groups are persistently viewed and treated as less capable or deserving. According to her, a group is unjustly subordinated when four conditions hold: (i) its members have less social and political power and de facto authority across multiple contexts; (ii) they possess or are ascribed traits that attract less consideration or greater censure than similar traits in dominant groups; (iii) those traits are shaped by or reinforce stereotypes that rationalize their lower status and treatment; and (iv) society is structured in ways—through structural accommodations—that tacitly prioritize the interests, values, and experiences of dominant groups while overlooking or minimizing those of marginalized ones. In contrast to Hellman's focus on expressive harm, Moreau shows that discrimination can be wrongful by entrenching social inequality through structural means.

In order to frame wrongful discrimination as unjust, we need to understand how subordination and demeaning treatment violate principles of a just society put forth by Elizabeth Anderson through democratic equality in her paper, "What is the Point of Equality". Anderson argues that justice must be grounded in democratic equality, a relational egalitarian theory that views equality not merely as a distribution of goods but as a structure of social relations. Democratic equality aims to dismantle the systems of stigma, hierarchy, dependency, and exclusion that prevent people from functioning as equal members of society. Unlike luck egalitarianism, which ties justice to compensating people for brute luck while holding them responsible for their choices, democratic equality focuses on ensuring that all individuals enjoy equal moral and social standing. Its core aim is to secure the social conditions necessary for freedom—that is, the ability of all individuals to participate in a cooperative society as respected equals. In Anderson's view, justice requires that people not be forced to degrade themselves to receive aid, nor marked as inferior because of what they lack. Instead, society must be organized so that no one is dominated or excluded from full citizenship. Equal respect and participation are non-negotiable features of a just society, not outcomes to be earned or redistributed post hoc, only after misfortune strikes.

I would argue that at the core of Anderson's theory of democratic equality is substantive equality of opportunity—a more demanding standard than formal or meritocratic views. Substantive equality requires not just formal access but the real, material means to compete on fair terms. Anderson adopts this insight, arguing that justice requires access to education, healthcare, and social recognition so that all individuals can develop the capabilities necessary for equal citizenship. Substantive opportunity thus depends not only on distributive fairness but also on social respect. Equal moral regard is not merely the outcome of opportunity—it is a condition for having one. Because democratic equality demands that individuals stand as equals in a cooperative society, substantive equality of opportunity is not just compatible with justice—it is a core requirement of it. This framework helps us understand why demeaning and subordination render discrimination unjust. Since substantive equality of opportunity is central to democratic equality, and democratic equality is a condition of justice, it follows that policies or technologies that subordinate or demean are not only wrong—they are unjust.

In 2014, Amazon developed an experimental AI-driven hiring tool aimed at streamlining the recruitment process by automating resume evaluations. The system was designed to rate job applicants on a scale from one to five stars, emulating the company's product review system. To train the AI, Amazon utilized a decade's worth of resumes from previous applicants, predominantly from male candidates, reflecting the tech industry's gender imbalance at the time. Consequently, the AI learned to favor resumes that mirrored the male-dominated data, systematically downgrading those that included terms like "women's," as seen in phrases such as "women's chess club captain," or references to all-women's colleges. Does the Amazon hiring algorithm subordinate women? Following Hellman's criteria, we can evaluate this in two parts. First, does the classification express a strong form of disrespect, specifically, a denial of equal moral worth? I would argue that it does. The system effectively communicates that experiences tied to being a woman, such as participating in women's groups or attending historically women's colleges, are not seen as professionally valuable. It implies that women's credentials are less legitimate, or less worthy of recognition, simply because they are gendered. That message, even if unintentional, reinforces a hierarchy in which male-associated experiences are normalized and female-associated ones are marginalized.

Second, does this expression come from an actor with sufficient power or status to make it socially meaningful? AI systems are tools, not moral agents in the traditional sense. But as tools widely deployed by powerful institutions—and relied upon by decision-makers who treat their outputs as authoritative—they function effectively as agents in practice. When it filters out female candidates before hiring managers even see their resumes, it materially alters their chances of employment. That kind of exclusion shapes people's standing in the world. It prevents access to opportunities that confer recognition, income, and social inclusion. In that sense, the system's actions meet both conditions for wrongful discrimination under Hellman's theory: they express unequal moral worth, and they do so with enough institutional force to impose that meaning on others.

Does the Amazon hiring algorithm subordinate women? According to Sophia Moreau's framework, a policy or practice unjustly subordinates a group when four conditions are met. First, the AI system accords women less social and institutional power by systematically filtering out resumes that include terms like "women's" or reference all-women's colleges. This downgrades experiences that are meaningful markers of leadership and achievement for female applicants, while elevating experiences more common among men, thereby diminishing women's de facto authority and visibility within the hiring process. Second, the traits associated with women's resumes, such as participation in women-centered organizations or attendance at historically women's colleges, receive less consideration than similar traits associated with dominant groups. Third, this differential treatment reinforces broader stereotypes: that women's achievements are less legitimate, or that leadership within women's spaces does not translate to professional competence. Finally, this system exemplifies a structural accommodation that privileges male applicants. It was trained on resumes from a male-dominated workforce and failed to correct for the gender imbalance in historical hiring. These four conditions—diminished power, asymmetrical consideration, reinforcement of stereotypes, and structural exclusion—together show how the Amazon AI system directly discriminates against women.

Now let's look at a hypothetical scenario that is a bit different and trickier because of how the model is developed. Imagine a tech company that uses an AI-powered resume screening system to assist in hiring decisions. This AI system is trained on historical data from previous "successful" employees, most of whom were white men. The model does not use race or gender as explicit input features. Instead, it evaluates resumes based on attributes such as the college attended, undergraduate major, GPA, work history continuity, and relevance of listed technical projects. These features are treated as neutral indicators of qualification, but they reflect patterns shaped by longstanding social inequalities. For example, applicants who attended historically Black colleges and universities (HBCUs), took time off due to caregiving responsibilities, or gained technical experience through non-traditional paths may be scored differently from those whose resumes more closely match the training data. Because the model has encountered relatively few examples of successful Black women in its training set, it may assign lower or more uncertain predictions to such candidates, without any explicit awareness of their race or gender. These uncertainty scores, however, are not conveyed to the hiring manager, who treats all recommendations as equally confident. As a result, applicants flagged with low certainty may be deprioritized or overlooked. On the surface, the system simply selects the "most qualified" candidates. But the composition of those selected reflects a pattern: a disproportionate number of white men advancing, and Black women being excluded at higher rates.

Some might argue that this hypothetical scenario does not rise to the level of wrongful discrimination, particularly under Hellman's account, because the model does not express a denial of equal moral worth. From Hellman's perspective, this case would likely not meet the threshold for wrongful discrimination because it fails to satisfy her first condition: that the action expresses a denial of equal moral worth. The model does not explicitly classify applicants by race or gender, nor does it treat these traits as disqualifying. Rather, it operates through proxy variables like education and experience. The exclusion it produces may be inequitable or damaging, but it does not communicate that Black women are morally inferior—it merely fails to correct for a background structure that already puts them at a disadvantage.

However, under Moreau's perspective, this scenario would constitute a case of wrongful indirect discrimination grounded in unjust subordination. The AI system in this scenario acts as a structural accommodation that privileges white male applicants over Black women. It does this by favoring criteria, such as elite university degrees, continuous internship experience, and certain technical project credentials, that are more readily accessible to white men due to historical and social advantages in STEM education and hiring. The four conditions of subordination are all satisfied: diminished power, asymmetrical consideration, reinforcement of stereotypes, and structural exclusion. While the system may not explicitly reference race or gender, the disproportionately low selection rate of Black women exposes the structural nature of the exclusion. What appears to be a neutral process in fact reproduces and legitimates social disadvantage through indirect but powerful means. Thus, according to Moreau's framework, this scenario constitutes wrongful discrimination because it causally sustains unjust subordination.

To conclude, this paper has addressed the question of whether it is unjust when AI systems disproportionately select men, especially white men, over women and Black women in particular. Through the analysis of one real case involving Amazon's resume-screening tool and a hypothetical example of biased AI-based resume ranking, I have shown that such disparities can constitute wrongful discrimination. In the Amazon case, the system directly demeaned and subordinated women, clearly violating their equal moral worth and reinforcing group-based disadvantage. In the hypothetical case, although the system did not explicitly demean, it indirectly subordinated Black women through structural accommodations and biased proxies that systematically disadvantaged them. As I argued, both demeaning and subordination violate the conditions of substantive equality of opportunity, which is a core pillar of Anderson's theory of democratic equality. Since democratic equality defines the moral foundation of a just society under relational egalitarianism, it follows that AI systems that disproportionately select white men over women, especially Black women, are not only discriminatory but also unjust.

In this paper, I will argue that surveillance capitalism diminishes autonomy. I will first define surveillance capitalism. Then, I will propose criteria that diminishes autonomy. Then I will argue that surveillance capitalism diminishes autonomy by analyzing Cambridge Analytica's scandal. Then I will provide an objection to the claim and a rebuttal.

Using large amounts of data or information to alter behavior is coined as "surveillance capitalism" by Shoshanna Zuboff. More formally, surveillance capitalism is defined as "the unilateral claiming of private human experience as free raw material for translation into behavioral data." This data is then "computed and packaged as prediction products and sold into behavioral futures markets -business customers with a commercial interest in knowing what we will do now, soon, and later." Surveillance capitalism has the following foundational components: behavioral surplus, machine intelligence, economies of scale, and action. Behavior surplus is considered exhaust or excess data that is gathered but not utilized to satisfy the consumers' needs. It is the kind of information that is used to claim private human experience, data that is not required for product and service improvement, but rather data that can be translated to behavioral data to produce predictions. What is important to note is that we are not consensually providing this data, it is collected without our knowledge, or permission. Machine intelligence is utilizing machine learning algorithms, such as neural networks, to create prediction or classification models based on relevant data and behavioral surplus. Economies of scale are the consequence of automated systems that track, hunt, and induce more behavioral surplus. The greater the data the better the predictions, which is why data must be acquired at a large scale. Economies of action utilizes behavioral surplus and machine learning to not only predict behavior, but alter an individual, group, and population's behaviors.

To determine whether surveillance capitalism diminishes autonomy we must determine what autonomy entails. Having autonomy is a spectrum, where some agents can have more or less autonomy than other agents, depending on their circumstances. I will not be arguing the case that surveillance capitalism entails no autonomy, but rather that it further diminishes one's autonomy. I would argue that an agent's autonomy is further diminished if their decisions and actions are unknowingly manipulated by direct forces.

Manipulation and influence are both tactics of persuasion however autonomy diminishes due to manipulation, which has three conditions. The first condition is that the primary intent of a manipulator is to reach their own desired outcome. Other agents are essentially pawns that allow them to achieve this outcome. There is a procedural difference in manipulation over influence. Manipulation primarily considers the manipulator's own interest, subsequently devising strategies to align another individual's goals, beliefs, and desires in a way that ultimately serves these interests. In contrast, influence primarily takes into consideration an agent's goals, beliefs, and desires and then determines possible decisions and actions that would align with and potentially benefit the agent's best interest. Therefore manipulation's primary intent is to guide agents to follow the manipulator's best interest rather than the agent's best interest. This means that the agent has no other options except for the option supplied by the manipulator. The second condition is that manipulation capitalizes on fear and insecurity to hinder rational decision-making so that the outcome that the manipulator desires is achieved. Rational decision-making is a tricky subject, however, what is meant by this notion is the ability for an agent to reflect on their goals, beliefs, and desires to come to a decision. This is hindered when persuasion capitalizes on fear and insecurity. Without the ability to reflect, agents will make quick and rash decisions that will not be from a place that is in line with their goals, beliefs, and desires but rather from a place of anxiety and apprehension. Persuasion through influence does not use a fear-based approach but rather a procedural approach that outlines options and consequences. Influence has reasoning baked into the tactic itself. Rather than a fear-based approach influencing matches decisions to goals allowing a person to carefully and thoughtfully self reflect, allowing for rational decision making. The third condition is that manipulation is pervasive and invasive, giving an agent no choice. It is not enough for one statement to be considered manipulative but rather it is the consistent proliferation of fear-induced statements until the desired behavioral change is achieved. While agents often seek out influencers, manipulators enter into the lives of agents to achieve their goals.

Autonomy is not simply diminished by manipulation because manipulation happens even in cases of which we are aware. In those instances, we can often interfere and redirect our beliefs and desires to align with our self interest. The diminishment of autonomy occurs when an agent is unaware they are being manipulated because they don't know how they are being manipulated hence lacking an option to stop the manipulation, and exit. Autonomy is also diminished by direct forces. These are forces that interact with a specific agent for a specific reason. The addition of direct forces is to evade the diminishment of autonomy by indirect forces. Indirect forces are societal structures such as home environment, socioeconomic background, race, sex, education, government, etc. These structures, while they do focus on specific agents, don't interact with the agents through targeted communication. Further, these structures are long-standing and are arguably baked into our identity and decision-making. Therefore, indirect forces are not external to ourselves but rather internalized subconsciously because of how one grows up and traverses the world based on these forces. I am not arguing that societal structures don't diminish one's autonomy, in fact, I think they do. However to determine whether surveillance capitalism further diminishes one's autonomy we cannot look at structures that already threaten and diminish our autonomy.

Now that the criteria for diminishment of autonomy has been proposed, I can determine how surveillance capitalism fulfills this criterion. To do so I will look at an example of a company that utilized surveillance capitalism. Cambridge Analytica, or CA for short, was a UK consulting firm owned by the billionaire and Donald Trump backer Robert Mercer and run by CEO, Alexander Nix. The firm was responsible for performing data analysis services for Ted Cruz's and Donald Trump's presidential campaigns. CA utilized a tremendous amount of nonconsensual data to perform data analysis services for their client. According to Chris Wylie, a whistleblower who formerly worked for CA, professor Alexander Kogan paid approximately 270,000 people to take a personality quiz. However, to be paid for their survey, users were required to log into an app that harvested as much data as it could about the user. It provided personally identifiable information such as real name, location, and contact details as well as access to the user's Facebook profiles. The app did the same thing for all the friends of the user who installed it but were not paid to fill out the survey. 270,000 people quickly amassed 87 million users. The users were unaware of the extrapolation of all the behavioral surplus data. Nix claimed to have the data resolved, "to an individual level where we have somewhere close to four or five thousand data points on every adult in the United States." The firm then built 253 predictive algorithms based on 253 different machine-learning approaches. Based on the data and the prediction models each voter was assigned a score based on 5 personality traits: openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism. Ads were then designed to target particular personality traits, for instance, neurotic people were shown warfare. The goal was to identify triggers that people have to move them from one state to another. These triggers were rooted in insecurities and fears. Roger McNamee, an early Facebook investor states that "they created a set of tools to allow advertisers to exploit that emotional audience with individual targets, manipulating users' realities." However, the bulk of ad resources did not go to the entire population but were rather targeted to the malleable minds. They called these voters the "persuadables" and determined them based on personality quizzes and Facebook dataset. According to Brittany Kaiser, the former business development director for CA, "We bombarded them [the persuadables] through blogs, websites, articles, videos, ads, every platform you can imagine until they saw the world the way we wanted them to. Until they voted for our candidate". This was an important demographic to target because in the US only about 70,000 votes in the three states, the swing states, decided the election.

CA's methods are a direct manifestation of surveillance capitalism. By scraping data non-consensually CA claimed the private human experience and translated it for behavioral data. By creating predictions of personality archetypes, CA manufactured ads that matched those archetypes and sold them to campaigners, and business customers with a commercial interest in knowing what we will do now, soon, and later. Economies of scale were illustrated by harvesting the data of 87 million users. Economies of actions were illustrated by conditioning and distorting the behavior of "persuadables" through pervasive targeted ads. Surveillance capitalism in the context of CA, manipulated agents, specifically the "persuadables". CA's wanted to help their clients win elections. The intention here is clear because the company desired the profit of the presidential candidates they hired. Therefore their purpose was to bolster their client's positions amongst the people. CA's intentions were not in support of doing what is best for the people based on their personality but rather to make ads that utilized their personality to best distort their desires and beliefs, so they could align with CA's desired outcomes. By characterizing users into different personalities, CA produced ads that targeted the fears of each personality type. CA capitalized on the fears and insecurities of people to trigger people into voting for their client. People were also shown ads constantly making the fear-motivated display invasive and persistent. Ads would pop up every time people used the internet, making the manipulation inescapable. Users were also completely unaware that the ads they were being shown were purposefully targeting their greatest fears or insecurities because they were unaware that CA had their personal data. Further, CA is considered a direct force because it was composed of a group of agents that chose which specific users to interact with, "persuadables", to trigger behavioral changes so that they would vote for the republican incumbent. Therefore surveillance capitalism as shown by CA diminished autonomy due to the unknowing manipulation by direct forces.

One could argue that surveillance capitalism is improving customer experience and satisfaction and is just a better marketing strategy because the methods of ad targeting are more efficient and refined. Most companies, before surveillance capitalism, utilized marketing methods that were meant to persuade individuals to buy their products. For companies to make a profit, they had to create tools that convinced consumers. Surveillance capitalism can hone what the people want better, through behavioral surplus, so the ads produced are more effective. Therefore even if we posit that ads created by companies in the past diminished our autonomy, surveillance capitalism wouldn't diminish it any further, because it is just a way of bettering predictions and tailoring persuasive content to a group of consumers. However, I would argue that surveillance capitalism does diminish autonomy even further for the very fact that it utilizes behavior surplus data which in the past was not accessible to companies. The distinguishing factor between past capitalism and current surveillance capitalism is that tech companies can gather personal experience without our knowledge and use it to make us puppets in profit-driven games. Before utilizing behavioral surplus, companies did make ads based on fear and insecurities but they were unable to pinpoint where to distribute such ads which made it far less invasive. Surveillance capitalism, because of its precision, can implement repetitive conditioning that manipulates behaviors in a revolutionary and profound manner. While surveillance capitalism can improve customer experience in terms of better product recommendations, it has also been utilized as a weapon for altering the democratic process. The question then becomes what is more important, upholding a better customer experience or salvaging and maintaining what little fragments remain of our autonomy.

Other possible objections could be made such as autonomy is only diminished when the unaware manipulation by a direct force is successful. However, this brings to light a deeper psychological question of how susceptible we are to manipulation. Further one could also argue how our autonomy could be diminished if we never had any autonomy to begin with under our current system? These arguments hold merit but are beyond the scope of this paper. Surveillance capitalism is a pressing issue on how our decision-making skills will be affected. As machine learning algorithms progress and more data is harvested, better predictions will be made about users, utilizing more robust personality descriptions. If algorithms are able to successfully pinpoint what fears and insecurities we have and trigger us based on those fears constantly, autonomy will be diminished. Our decisions, desires, beliefs, and goals will no longer be ours but lie at the discretion of profit-hungry manipulators.

Thomas Kuhn proposes what constitutes a scientific revolution in his book "The Structures of Scientific Revolutions". He applies his framework onto historical examples such as the chemical revolution, and the quantum theory of light. In this paper, I will be exploring whether theories that have been called revolutionary are actually revolutionary according to Kuhn's framework. I will focus on the advent of artificial neural networks. I will first reconstruct the Kuhnian framework of scientific revolutions. Then, I will briefly describe the major points in history that led to the creation of artificial neural networks. Then, I will contextualize that history with the existing paradigm. Then, I will discuss whether artificial neural networks are considered revolutionary in the domain of cognitive science according to Kuhn.

Normal science is almost analytically associated with a paradigm. A paradigm is a concrete proposal that is convincing for a certain phenomenon—"it is an object for further articulation and specification under new or more stringent conditions" (Kuhn, 23). Paradigm-based research has three focal points: a class of facts that increase the accuracy and scope of the paradigm, a class of facts that can be compared directly with predictions, and the further articulation of the paradigm. Normal science does not aim to invent new theories but rather to articulate the theories the paradigm already supplies.

Scientific revolutions occur when normal science breaks down in light of a crisis—"the persistent failure of the puzzles of normal science to come out as they should" (Kuhn, 68). The rejection of a paradigm only occurs when there is an alternative candidate that is incompatible with the existing one—"scientific revolutions are here taken to be those non-cumulative developmental episodes in which an older paradigm is replaced in whole or in part by an incompatible new one" (Kuhn, 92).

Artificial neural networks have been considered by many to be revolutionary. In 1943, McCulloch and Pitts described how neurons might work by modeling a simple neural network. In 1957, the Perceptron was demonstrated by Rosenblatt. In 1987, Rumelhart and McClelland published "Parallel Distributed Processing" which developed more complex systems and formalized connectionism.

During the 1940s the dominating paradigm explaining how the mind works was behaviorism. It stumbled upon various difficulties and induced a crisis. In 1956, cognitive science was pioneered. In 1967, Putnam introduced the classical computational theory of mind (CCTM), which stated that the human mind is an information-processing system and mental processes can be understood as Turing-style computation. For a while, CCTM was the dominating paradigm. In the 1980s connectionism became popular, challenging symbolic computation by proposing that cognitive processes could be understood as interactions within networks of simple units like neurons.

For ANNs to be considered revolutionary, we would have to determine whether connectionism caused a paradigm shift. The main crisis seemed to be that CCTM was not "biologically plausible"—its theoretical aspects did not match nature. Rumelhart and McClelland believed that CCTM could not account for perception and memory because it relied on Turing-style computation. It does seem that connectionism and CCTM are incompatible theories. However, it does not seem to be the case that connectionism is more widely accepted or that it better explains cognitive processes than CCTM. In 1988 Fodor and Pylyshyn argued that "systematicity" and "productivity" fail in connectionist models. Some researchers conclude CCTM is better suited; others argue connectionism is more biologically plausible.

Following the Kuhnian framework it would seem plausible to argue that it is too early to deem ANNs revolutionary within cognitive science. There has been no paradigm shift since connectionism fails key aspects that CCTM succeeds in, while CCTM fails in key aspects that connectionism succeeds in. Both theories have followers performing normal science. While it could be the case that ANNs end up being revolutionary it would seem that scientific revolutions are not perceptible until they have already occurred and much time has passed. Therefore ANNs ought not to be considered revolutionary because there either is still considerable trust in CCTM or while CCTM may be gaining distrust, not enough time has passed for us to notice widespread distrust and a new willingness to accept connectionism over CCTM.

This paper explores whether Karl Popper's methodology of testing theories survives in modern-day scientific practice of testing theories in the domain of data science. In this paper, I will first lay out Popper's methodology of testing theories through falsification. Then I will lay out the data science methodology of testing theories through hypothesis testing. Then I will provide an example theory and implement both frameworks. Then I will highlight the similarities between both frameworks. Then I will argue how Popper's falsification methodology is not used in modern day data science's hypothesis testing methodology.

Popper's framework of falsification follows the notion that in order to falsify a theory, something has to contradict the theory. First, the theory must be a universal statement, a statement that claims to be true for all space and time, such as "all ravens are black." Second, a theory must be falsifiable. Basic statements have the logical form of singular existential statements. For instance, "there is a white raven in the space-time region k." A theory is falsifiable only if it divides the class of all possible basic statements into two non-empty subclasses: the class of basic statements which are inconsistent with the theory (potential falsifiers), and the class it does not contradict. The guiding principle is that a theory is falsified if, and only if, we have accepted basic statements which contradict it. Popper suggests coming up with a low-level empirical hypothesis. For instance, to falsify "all ravens are black" we could come up with "there is a family of white ravens in a zoo in New York." If this lower level hypothesis is corroborated by the acceptance of basic statements, then it serves as the negation of the theory, and the theory is falsified. However if not corroborated then the universal theory is corroborated since it survived the severe test. It is important to note the asymmetry between falsification and corroboration.

In data science, theories are tested through hypothesis testing. The alternative hypothesis is a claim about the theory: "playing chess affects critical thinking skills." The null hypothesis is a statement about the treatment having no effect: "playing chess does not affect critical thinking skills." We then find the probability of the data occurring given the null hypothesis, called the p-value. If we find that the probability is very low, below the conventional significance level of 0.05, then we say the p-value is significant and the data contradicts the assumption that the null is true. If the data is greater than the p-value then we fail to reject the null. Note this method does not make any claims about accepting or rejecting the alternative hypothesis.

To test the theory "the drug NZT affects IQ" under Popper's framework: the lower level falsifying hypothesis would be "there is a group of undergrads where NZT does not affect IQ." If basic statements corroborate this, the theory is falsified. Under hypothesis testing: the null is "NZT does not affect IQ." If the p-value is 0.01 (less than 0.05), we reject the null. Note that we just make a statement about the null, not the alternative.

It would seem plausible that data maps onto basic statements, and that universal and alternative hypotheses are equivalent since they share the same logical form. However, I would argue that the lower level empirical hypothesis and the null hypothesis differ in logical form. The lower level hypothesis is a pure existential statement, but the null hypothesis has the logical form of a universal statement. This means it cannot necessarily negate the alternative hypothesis.

The distinction lies in the decision step. Popper would suggest that if the basic statements corroborate the lower level hypothesis then we can contradict the universal hypothesis. However, in hypothesis testing we never test the data assuming the alternative hypothesis is true. We would never test the data given "NZT affects IQ" since we don't have data of what the population looks like when NZT affects IQ. Since we cannot test a lower level hypothesis of this kind we are not able to falsify nor corroborate the alternative hypothesis. This is where data science focuses more on falsifying null theories in order to give us good reason to believe the alternative theory. When the null is rejected we can increase our confidence in the alternative. This would not fit Popper's framework because falsifying one theory does not indicate means for corroborating another theory. Since data science does not attempt to falsify the alternative it would seem that data science stands as a counter example to how Popper proposes science should be done.

Determining what race is metaphysically has been a controversial debate amongst many accredited philosophers. Race is a phenomenon unlike any other, that has very serious implications on not only the way people live but also the way people identify themselves. Knowing our identity is crucial as it tethers us to reality and our place in the world. Race also lends a direct hand to racism, which has plagued our country for centuries. It may come as a surprise that race, something that has affected almost everyone, has no agreed-upon metaphysical meaning. In this paper, I will explore the constructionist viewpoint of race proposed by Sally Haslanger.

Sally Haslanger proposes that race is a socially constructed concept and therefore should be regarded with concepts that common people use to categorize races. In contrast to eliminativism and naturalism which demand that races should be considered natural kinds, Haslanger suggests that races are social kinds. By claiming that races are social kinds Haslanger proposes that race ought not to be determined by experts but rather how common people perceive race, the "folk notion". Therefore Haslanger defines races to be "observed or imagined bodily features presumed to be evidence of ancestral links to a certain geographical region."

From a constructionist viewpoint, features are categorized based on what people think is an important distinction to make when trying to push a social agenda. For instance determining the race Black during slavery depended on skin color because in invoking you vs them mentality between slave owners and slaves, as well as justifying heinous treatment, it was needed that a plain sighted difference in features could be attributed to make-believe bad characteristics. Haslanger's definition of race is nuanced to the extent where race itself is correlated to phenomena of oppression. The constructionists view objects eliminativist and naturalist views because their arguments rely on whether biological variations can be classified into specific genetic groups. Haslanger retorts that it is not a matter of whether the metaphysical possibility of races being biologically real is significant, but whether the purpose of classifying them as such is significant.

Race is a phenomenon that can be said to work like a recipe: first a group of people must decide what phenotypic trait is significant. Second, each person with the significant phenotypic trait is considered a race. Third, trait x is given attributes, usually negative, to correlate with behavioral characteristics. Fourth, each person who has trait X is grouped with these characteristics and subordinated because of them and therefore racialized. Haslanger's definition however does not exactly indicate who is placed in what category. Haslanger would retort that because race is a sociological phenomenon, who fits into which racial category is highly dependent on our identity—cultivated through experiences.

Haslanger's view that race is socially constructed provides us the best metaphysical description of what race is. This metaphysical description allows us to define race in a way that is consistent in propelling our fight for racial justice. Therefore employing a definition that may have explanatory gaps and may not be fully metaphysically correct in the eyes of other philosophers is still our best method of granting others a definition of race that can be accepted and used for policy and social reform.

The ongoing debate about what exactly constitutes gender has been circulating for years. Since many argue that gender is a socially constructed term, not on the basis of natural sciences, coming up with criteria for categorizing one as woman, man, or other, has been largely contested. Philosophers have taken an ameliorative inquiry into defining gender, which lies in developing a target concept that aims to determine the shared meaning that a particular group should aim to be the dominant ideology. In this paper, I will explain Haslanger's definition of gender as a class. Then I will propose objections to Haslanger's definitions using Katharine Jenkins's four scenarios. Then I will propose Jenkins's ameliorative definition of gender.

Sally Haslanger proposes that "S is a woman if S is systematically subordinated along some dimension, and S is 'marked' as a target for this treatment by observed or imagined bodily features presumed to be evidence of a female's biological role in reproduction." She develops three conditions. The first condition aims to say that what gender one is a part of depends on bodily features that are either observed or imagined to be an indication of one's role in reproduction. The second condition states that given condition one, a person becomes a part of the dominant ideology—the social position rooted in subordination and domination. The third condition states that to be gendered as a woman it is necessary to be classed as a woman.

Katharine Jenkins responds to Haslanger's definition in a series of scenarios where trans women are excluded. In Scenario 1, a trans woman does not publicly present as a woman and is perceived as a man. By Haslanger's definition she fails to be perceived as a woman. In Scenario 2, a trans woman publicly presents as a woman but her gender presentation is not respected. She is seen as a man "pretending" to be a woman. In Scenario 3, a trans woman publicly presents as a woman and her gender presentation is respected because she is perceived as having associated bodily features. No conditions would be violated. In Scenario 4, a trans woman publicly presents as a woman and her gender presentation is respected unconditionally, her body parts are not taken into account. This would violate Haslanger's account because there is no way of determining who is a woman without acknowledging reproductive organs.

Jenkins believes that being classed as a woman and gender identity should be considered in conjunction with one another. She proposes a definition that combines gender as a class and gender as an identity. Gender as a class holds the same definition that Haslanger proposes, while gender identity uses an internal map: "If someone has an internal map that guides them in the same way as other women who are classed as women are guided then they are also considered women based on self-identification." Jenkins's concept of gender provides autonomy in being a certain gender. However, it is worth noting that Jenkins mentions that to be a woman, the gender identity must be somewhat attuned to the norms of society. Jenkins's account of gender strives to be neither emancipatory nor oppressive but rather guides a person through the social and material reality of the world.

I believe that a more plausible future goal lies in developing legislation and altering the conversation around those who subordinate on the basis of gender. Engineering a new concept of gender, while a noble goal would be extremely difficult to employ. However, employing this view of gender in terms of inclusivity is extremely emancipatory for women to act as they desire while also securing a part of their identity.

The following argument concludes that Aristotle's ethical theory is not a plausible ethical theory. Premise 1 follows that any ethical theory must adhere to principles of sacrifice and motivation. Sacrifice being that in some cases, we morally ought to do actions that are not good for us. Motivation being that in some cases, the considerations that ought to motivate us to act have nothing to do with what is good for us. Instead, in such situations, we should act entirely for the sake of others. In this essay, I will first assess premise 2 by providing examples from the Nichomachean ethics that illustrate sacrifice. Then I will provide examples that do not illustrate motivation, arguing that premise 2 is true. Then I will assess premise 1 and attempt to prove it false. Through the rejection of premise 1 and the acceptance of premise 2, I will conclude that the objection to Aristotle's ethical theory, while valid, the argument fails to be sound.

I believe that Aristotle's ethical theory does follow sacrifice. Aristotle's ethical theory stems from the idea that the finest good is happiness but happiness is activity. Aristotle argues that living well is the activity of the soul in accordance with character virtues. In terms of sacrifice, Aristotle talks about the actions of an excellent person as character virtue. Character virtues are states that do not overindulge or under indulge. Habituating bravery requires practicing actions that are frightening in that they are life-threatening actions. Actions required in habituating bravery may not be good even though the action produces the finer good of acquiring a character virtue. This is an example where Aristotle's ethical theory shows that we morally ought to do actions that are not good for us in order to achieve happiness.

I agree with premise 2 in that Aristotle does not follow motivation in his ethical theory. Aristotle says "to a friend, however, it is said, you must wish for good for his own sake." This claim does not necessarily indicate that an excellent person will wish for good only for the sake of their friend because Aristotle further claims that good friendships have reciprocated goodwill. An excellent person in a good friendship acts also for the sake of acquiring the virtue of justice. "For a recipient of a benefit returns the goodwill for what he has received, thereby doing the just thing." An excellent friend, therefore, does not solely do actions for the sake of their friend but also to practice justice. Aristotle states: "the excellent person labors for his friends and for his native country, even if he ought to die for them if he must; for he will sacrifice money, honors, and contest goods altogether, achieving the fine for himself." Every action that the excellent person does, even for others such as sacrificing, will be the motivation of acquiring happiness. Therefore it is not possible for the excellent person to act entirely for the sake of another in Aristotle's theory.

Now I will assess the truth of premise 1. For sacrifice: in Jain culture, fasting is observed for the purpose of achieving Nirvana. The action of fasting can be considered excruciating for most. While the action in and of itself is not good and painful we morally ought to prescribe to the action in order to achieve the virtue of asceticism. Another case is fighting in a war to acquire the finer good of bravery. For motivation: I will argue that the definition of motivation is in fact impossible for one to achieve. Every action that an excellent person takes is to become or maintain their excellence through habituating character virtues in order to live happily. A tricky case is motherhood—however the excellent mother would wish not only for the finer good of being a good mother but also for the finest good of living well. There seems to exist no such case where the considerations that ought to motivate us to act have nothing to do with what is good for us.

For these reasons, while premise 2 seems to hold true, premise one seems to be false since the motivation proposed does not seem to be possible. Therefore it need not be the case that motivation needs to be part of an ethical theory. Since both premises of the argument are not true while the argument can be considered valid it is not sound. Therefore the conclusion that Aristotle's ethical theory is not a plausible ethical theory does not follow from the argument.

In this paper I will outline Rousseau's argument that a group of people is a sovereign if the collective decisions are guided by the general will and if a group is considered a sovereign then each individual within that group is free. Rousseau's argument follows: If decisions are guided by the general will then everyone's private interests are respected. If everyone's private interests are respected then everyone has autonomy and self determination, or moral freedom. A group is considered sovereign if they have moral freedom. Then a group must be guided by the general will to be considered a sovereign. A sovereign entails civil freedom. One is free if they have moral and civil freedom. If a group of people are considered a sovereign then each individual within that group is considered free.

Rousseau claims that in order to be free you must have civil and moral freedom. In the state of nature humans are rational beings, however when there is scarcity, individuals have to depend on others property for self preservation. The importance of being able to preserve one's self without becoming dangerously dependent on another is an essential reason as to why Rousseau believes that a person must join a state, considered an association. The absence of dependency in conjunction with the ability to preserve one's self is what Rousseau considers civil freedom. While civil freedom is necessary it is not sufficient to be truly free in an association. Once one joins an association that will protect against dependencies, they must also make sure that the association allows for them to obey their own will. Self-governance refers to moral freedom which grants a member the right to make decisions based on their own interests.

An implicit premise seems to be that if a group of people are morally free then they are a sovereign, and moral freedom is only achieved through the general will. The sovereign is a collective that is guided by the general will. Rousseau distinguishes between the particular will, the corporate will, and the general will which takes into account the interests of the entire society. The general will is just the harmonization of particular interest which leads to the common interest. The general will can only be established through collective deliberation of the people through the democratic process. Rousseau claims that the general will can never err therefore if the general will is produced then a group of people must be sovereign. Everyone has the general will within them.

When a group is guided by the general will one gains moral freedom since they have the autonomy to govern themselves. One also gains civil freedom because if individuals have authority over decisions and are part of an association that guarantees self preservation, dangerous dependency is diminished. Rousseau explains that when one joins a sovereign they renounce their property to be redistributed so that one with more property can't enslave one with less.

A group is considered a sovereign when it is guided by the general will. An individual is only morally free if they are guided by the general will. The general will requires collective deliberation through the democratic process. If one is only morally free when guided by the general will, then how would they have the ability to deliberate collectively and implement the general will in the first place? Rousseau could answer that there is a distinction between collective moral freedom and individual moral freedom. A group would have to have collective moral freedom prior to deliberating on the general will. However when a collective is guided by the general will, individuals gain moral freedom. However it could be argued that a collective moral freedom can only be formed when individual moral freedom is guaranteed. The objection poses a notoriously perplexing dilemma: what came first, the chicken or the egg?

Cohen argues that wage laborers are collectively unfree but individually free in the institution of wage labor. Cohen defends this claim by objecting to both the Libertarian and Marxist argument regarding freedom within the institution of wage labor, or capitalism system. In this paper I will first reconstruct Cohen's objection to the Libertarian claim. Then I will reconstruct Cohen's objection to the Marxist claim. Through these objections I will show how Cohen constructs his claim regarding the freedom of wage workers. Then I will propose a possible objection to Cohen's view. Then I will provide a way that Cohen may respond to the view. Lastly, I will provide a plausible objection to Cohen's response.

On the Libertarian account of freedom, wage workers are free to sell their labor power. Cohen's objection aims to show that libertarians misuse the concept of freedom. Cohen begins by citing Flew, who defines libertarianism as "whole-hearted political and economic liberalism, opposed to any social or legal constraints on individual freedom." On this view libertarians believe that any social or legal interference constrains an individual's freedom (the neutral view). Capitalism rests upon private property. According to Cohen if capitalism rests upon private property then non-property owners will be unfree if interfered with. Cohen says there is a different concept of freedom that informs Libertarians: "I am unfree only when someone prevents me from doing what I have a right to do." This leads to the claim that protection of private property is not a restriction on anyone's freedom. However, Libertarians can only claim that A is unfree if they posit a neutral account of freedom. If they use both accounts they run into a dire contradiction.

Cohen's interpretation of the Marxist view is that "in capitalist society the great majority of people are forced to sell their labor power, because they do not own any means of production." Cohen argues that there are cases where proletarians have entered the bourgeoisie, therefore if anyone wanted to leave the proletariat class they could objectively. Cohen counters with the key analogy: imagine a room with two doors, two keys and ten people. One gets out and succeeds. Now there is one door, one key, and nine people remaining. Each of the nine is free to leave the room if they desire. This showcases that there is at least one exit that no one will attempt to use. However Cohen takes into consideration that one can only grab the key on the condition that others won't—"each is free only on condition that the others don't exercise their similarly conditional freedom." This contingency-based freedom is what Cohen coins collective unfreedom.

Cohen's theory lies on the assumption that wage workers have the possibility to exit the class. I would argue that there are some cases where it is virtually impossible for a wage laborer to leave the class. For instance Anita, a black woman that works in a factory. According to the Economic Policy Institute, until the 1970s employers' exclusion of black women from better-paying, higher-status jobs with mobility meant that they had little choice but to perform private domestic service work. For Anita it would be virtually impossible to escape. Cohen could respond that he is referring to "most" people. However, the key analogy can only be granted on the prodigious assumption that most proletarians are not trying to escape. What happens when six people in the room want the key, but two people grab it and both doors shut? In capitalism's essence there would need to be a substantial proletariat class therefore the key could not be reinstated for all six. Therefore most proletariat would not have the possibility to exit the class. It does not seem plausible that Cohen can grant individual freedom to wage laborers on a fickle psychological assumption.

A new regime of radical right-wing political ideology has begun to spread within the United States of America in the past few decades. While the growth of the political ideology was slow, careful nurturing by notable forefront donors like David Koch, Charles Koch, and Richard Mellon Scaife has come to have painfully significant effects. But how did these multibillion-dollar industrialists gain a seat at the allusive Supreme Court of the United States (SCOTUS) table?

On August 22th, 2022, an insider tip-off to the New York Times reported a 1.6 billion dollar donation from Chicago electronics manufacturing magnate Barre Seid to the Federalist Society, an organization led by infamous conservative libertarian lawyers dedicated to elevating conservative scholars to powerful judicial branch positions. This tremendous donation, while harmless at first glance, has contributed to affluent industrialists and conservative moguls clandestinely arranging for a foot in the door of the judicial branch. Donations to nonprofits and think tanks benefit from shielded donor disclosure and serve as a loophole for lobbying toward courts and political authorities without identification. The pernicious weapon of manipulation utilized to implement industrialist's goals in legislation has been coined as "Dark Money."

Dark Money has festered rampantly in the political arena. Mysterious donations supplied to select uber-conservative think tanks such as the Federalist Society, the Heritage Foundation, and Americans for Economic Prosperity propelled a wave of new academia and research surrounding conservative thought that has infiltrated many academic institutions, and, more perilously, law schools. The chairman of the Federalist Society, Leonard A. Leo, has motivated the careers of several conservative Supreme Court appointees, including Justices Brett Kavanaugh, Samuel A. Alito, Clarence Thomas, and Neil M. Gorsuch. By getting into the pockets of senators and the president, donors and nonprofits were able to stack the court with justices that are in line with industrialist and conservative think-tank values and missions. The federalist society came to dominate the supreme court by playing a long and arduous game, positioning their pieces on the board in the most precise manner before launching the final attack.

The effects of Dark Money within the Supreme Court have manifested themselves in the hard-hitting and monumental opinions of the last SCOTUS term. The Federalist Society adheres to an originalist interpretation of the constitution, and due to the lobbying of Supreme Court Justices using Dark Money a majority of the court also adheres to an originalist view of the Constitution. The instrumental effects have specifically materialized in the latest 6-3 opinions: West Virginia v. EPA (2022), Kennedy v. Bremerton (2022), and Dobbs v. Jackson Women Health Organization (2022).

In West Virginia v. EPA, the Supreme Court decided that the EPA cannot implement generational energy shifting on the basis of the major question doctrine. In Dobbs v. Jackson Women's Health Organization, the 6-3 decision overturned Roe v. Wade (1973) which guaranteed a constitutional right to abortions. In Kennedy v. Bremerton, conservative justices replaced the longstanding Lemon test with a strict interpretation of the Constitution.

The aforementioned court cases are merely three of the several cases argued and won during the last term in an originalist manner. Dark Money and its influence have infiltrated the Supreme Court and have a significant possibility to impose originalist rulings over the upcoming cases on the docket. Currently, the Johnson amendment is in place to regulate the influence of non-profits in politics, however, I believe that in order to truly regulate the influence of nonprofits in government, large-sum donations should be stifled. Specifically, donating to nonprofits that affiliate as think tanks should not be considered tax deductions for the donors. If legislation is not passed to limit donations and disclose donors, democracy could be thought of as a thing of the past as we enter a plutocratic form of government.