You are not my Type! - Archetypes among Grad Students - Agent-Based Modeling to Generative Social Intelligence

The Sticky Idea That Won’t Let Go

There’s something fascinating about ideas that stick with you—the ones that keep evolving, reshaping themselves across different stages of your life and career. For me, that idea has been generative social intelligence. It’s been with me for six years now, morphing from corporate consulting projects to hackathon experiments, and now into understanding the very archetypes of people around me in grad school.

“This one problem space has been stuck with me for the last six years on different stages, getting different shapes.”

The core fascination remains the same: How can we simulate human interactions and decision-making using data-driven methods? But the applications have been incredibly diverse—from predicting consumer behavior for beverage companies to building enterprise social networks, and now to understanding the complex ecosystem of graduate students pursuing degrees in computer science and data science.

The Corporate Genesis: Agent-Based Modeling in 2019

It all started back in 2019 when I was working on agent-based modeling for a consulting company. This wasn’t your typical “let’s build neural networks” project that everyone was jumping into. Instead, we were diving deep into something more fundamental: understanding humans, their decision-making, and their behavior through data science.

The project involved simulating thousands of individual agents to understand consumer behavior for a North American beverage company. We utilized market panel data and consumer panel data, building Bayesian networks to model individual decision-making processes. The question we were trying to answer was beautifully complex: What would be their behavior if the situations, offers, campaigns, or new products changed?

What made this different from typical data science work was its essence—we weren’t just building models to predict outcomes, we were trying to understand the why behind human choices. This philosophical approach to data science has stayed with me ever since.

The Evolution: Enterprise Social Networks and Synthetic Users

Fast forward to my time at a tech company, and the idea took on a new shape. I found myself often feeling isolated in the workplace environment, which got me thinking: Why can’t there be better social networks driven by agents?

This led to an interesting experiment during a company hackathon. While everyone else was building document editors and tool calling applications, I pitched something different: an enterprise social network with agents—a concept blending computer-human interaction with an ecosystem that lives alongside people.

The concept was elegant in its complexity:

• Individuals could create their own agents through detailed questionnaires
• We extracted user behavior patterns and personality matrices
• The system understood intent and what people expected from others
• It matched people for collaboration, hobbies, projects—finding companions for cycling, badminton, foosball, or work partnerships

This wasn’t just a dating engine; it was about making people find companions and collaborators to create genuine value together.

The Academic Intersection: Understanding Student Archetypes

Now, as a teaching assistant for quantitative methods in social sciences, everything has come full circle. The same framework that helped simulate consumer behavior and workplace connections is now helping me understand the diverse archetypes of graduate students around me.

The Data-Driven Discovery Process

Rather than relying on personal judgments or assumptions, I turned to social media discussions and forums where students openly share their frustrations, intents, and aspirations during their graduate school journey. The patterns that emerged were both predictable and surprising.

---

Based on the student profiles, several distinct patterns emerge around what frustrates students most and what they aspire to achieve.

Common Frustrations

The Theory-Reality Chasm dominates student concerns. Students consistently express feeling trapped between abstract coursework and the practical skills they need. They describe algorithms classes that feel disconnected from real applications, statistics courses that lack business context, and programming assignments that don’t resemble actual software development.

Siloed Learning represents another major source of frustration. Students pursuing healthcare technology find themselves in CS programs that barely mention medical applications. Those interested in finance struggle through data science courses that ignore financial modeling. The curriculum feels rigidly separated from the interdisciplinary reality of modern technology work.

Competitive Academic Toxicity creates stress for many students. The relentless focus on grades over genuine understanding, the pressure to maintain perfect GPAs, and the exhausting competition with peers from elite backgrounds leaves students feeling burned out before they even enter their careers.

Outdated Content frustrates students who see emerging technologies like blockchain, quantum computing, or AR/VR barely covered in their programs while the industry moves rapidly forward. They feel they’re learning yesterday’s tools for tomorrow’s jobs.

Limited Practical Experience leaves students feeling unprepared. They want more hands-on labs, real-world projects, industry partnerships, and opportunities to build actual systems rather than theoretical exercises.

Social Isolation and Imposter Syndrome particularly affects students from underrepresented backgrounds or those changing careers. They struggle to find mentorship, feel disconnected from peers, and question whether they belong in competitive programs.

Common Aspirations

Building Real Solutions drives most students. They want to create systems that actually solve problems, develop applications that people use, and work on projects that have tangible impact rather than academic exercises.

Domain Integration represents a core aspiration. Students dream of becoming the person who bridges technology with healthcare, finance, entertainment, or environmental science. They want to be domain experts who understand both the technical and business sides of problems.

Meaningful Work matters deeply to students. Whether through social impact technology, environmental solutions, educational innovation, or healthcare improvements, students want their careers to contribute to something larger than themselves.

Innovation and Entrepreneurship appeals to students who want to build new companies, create disruptive technologies, and shape the future rather than just work for existing organizations.

Research Excellence attracts those who want to push the boundaries of knowledge, publish groundbreaking papers, and contribute to fundamental advances in AI, computer science, or interdisciplinary fields.

Skill Mastery motivates students to become experts in emerging technologies, develop deep technical capabilities, and stay at the forefront of rapid technological change.

Work-Life Integration increasingly concerns students who want successful careers without sacrificing personal well-being, creativity, or other interests.

Mentorship and Community represents both an aspiration and a way to address current frustrations. Students want to find mentors for themselves while also becoming mentors for others, particularly those from underrepresented backgrounds.

The underlying tension appears to be between educational systems designed for broad foundational knowledge and student desires for specialized, practical, interdisciplinary preparation for rapidly evolving career paths. Students feel caught between institutional structures and personal ambitions, seeking ways to bridge this gap through projects, internships, and self-directed learning.

---

The Archetypes: A Living Taxonomy

Through this analysis, several distinct student personas emerged, each with their own unique blend of frustrations and motivations:

The Theory-Practice Bridge Builders

“There’s a huge disconnect between theoretical coursework and what the industry actually wants.”

These students are constantly frustrated by the gap between academic rigor and practical application. They’re taking analysis of algorithms or other heavily theoretical subjects but worry these aren’t giving them the edge to build projects or work on industry-relevant problems. They represent the most common archetype—students who came to graduate school expecting a bridge to industry but finding themselves on an academic island.

The Domain Integrators

These are the students asking: “I’m studying mechanical engineering, but where’s the AI in mechanical engineering?” They’re looking for interdisciplinary connections—healthcare applications of CS, AI in mechanical engineering, blockchain in finance. They want their technical skills to solve real-world problems in specific domains, not just exist in theoretical vacuum.

The Research-Oriented Traditionalists

Interestingly, there’s a counter-narrative here. Some PhD students and research-focused individuals are frustrated by the excessive obsessiveness towards startup building. They feel this entrepreneurial fever is reducing the quality of academic advising and research mentorship. They want deep, focused academic work but feel pressured by the startup culture.

The Overwhelmed Navigators

These students feel completely overwhelmed about career direction. They’re intellectually capable, often having rich conversations about mental health applications, children’s health, or educational frameworks, but they’re paralyzed by choices. They apply to jobs continuously but struggle to articulate their unique value proposition.

The GPA Maximizers vs. The Early Applicants

Two distinct tactical approaches emerged: students who focus intensely on achieving high GPAs while simultaneously applying for internships day and night, and those who believe in being early applicants with a clear policy of getting ahead of the curve. Both strategies reflect different philosophies about success in graduate school.

The Horizontalists

This is perhaps the archetype I most identify with—the horizontalists. These are students who approach problems as first priority and take efforts to learn, experiment, and connect different disciplines to solve particular problems. Rather than deep specialization in one area, they’re building broad, interconnected knowledge that can be applied across domains.

The Broader Implications: Beyond Student Life

What makes this framework powerful is its universal applicability. The same principles that help us understand graduate student archetypes can be applied to:

• Understanding consumer behavior in different markets
• Designing better employee engagement strategies
• Predicting community responses to new products or services
• Creating more effective educational curricula
• Building better social platforms and networking tools

The Technical Challenge: Scaling Social Intelligence

The computational complexity of simulating thousands of agents in parallel remains a significant challenge. During our earlier corporate work, scaling from R to cloud-based solutions required substantial AWS infrastructure and transitioning to more scalable languages. When you’re trying to simulate extended time periods—quarters or seasons with their various cultural moments—the compute requirements grow exponentially.

The Continuous Learning Loop

What excites me most about generative social intelligence is its Bayesian nature—it continuously updates beliefs and models based on new evidence. You start with simulated personas based on initial data, but over time, real-world signals refine and ground these synthetic agents in actual behavioral patterns.

This creates a virtuous cycle: synthetic data helps us understand probable behaviors, real-world feedback corrects our assumptions, and the model becomes more accurate and useful over time. It’s not just about prediction; it’s about creating a living, learning system that grows with the community it’s modeling.

The Human Element: Understanding Choice and Connection

At its heart, this work is about understanding the fundamental human questions: How do we make decisions? How do we choose universities, courses, friends, career paths? How do random people from different countries come together in graduate school and form tight-knit communities within weeks?

“I always try to watch how people really take decisions or what makes people take certain choices—how they choose universities, programs, how they make friends once they come into grad school.”

These aren’t just data science problems; they’re profoundly human questions that require both technical sophistication and genuine empathy.

Looking Forward: The Future of Social Intelligence

As I continue exploring this space, I’m seeing how generative social intelligence intersects with broader trends in AI and human-computer interaction. Companies are already using these approaches for market research, user experience design, and product development. The challenge is making these tools more accessible and more grounded in real human experience.

The goal isn’t to replace human insight with artificial simulation, but to augment our understanding of human behavior with data-driven models that can help us make better decisions about everything from course design to community building to product development.

---

This exploration continues to evolve, shaped by both academic rigor and practical application. If you’re interested in the intersection of behavioral science, data science, and social systems, I’d love to hear about your own experiences with understanding human archetypes in your domain.

What archetypes do you see in your own field or community? How might we better understand and serve these different personas?