00. RESEARCH & INTERESTS

HOW I THINK

Understanding the systems behind

human behavior, products,

and technology.

How do we understand the systems that shape human decisions?

My work and research begin with a persistent interest in what happens beneath the surface of products, interfaces, human behavior, technology, business, and society. Every interface is part of a product. Every product exists within a system. Every system influences human behavior, decisions, and relationships.

FIVE CONNECTING AREAS & ASSOCIATED RESEARCH QUESTIONS

01. AREA

PEOPLE

Perception, attention, cognitive limits, emotion, and identity in human judgment.

Why do people notice certain signals, overlook others, and rationalize choices?

02. AREA

PRODUCTS

How problem framing, workflows, interfaces, and defaults shape what users can do.

What separates a surface feature from a product that resolves a real problem?

03. AREA

SYSTEMS

How incentives, rules, dependencies, and feedback loops drive outcomes.

How do individual actions combine into systemic behavior no single person intended?

04. AREA

TECHNOLOGY

How software, automation, and AI shift from passive tools into active decision participants.

What happens to agency and accountability when software begins acting?

05. AREA

SOCIETY

How institutions, economics, media, and collective identity shape trust and empathy.

How do social, economic, and geopolitical forces shape everyday priorities?

VISUAL SEQUENCE OF INQUIRY

Observation → Pattern → Question → Investigation → Framework → Evidence → Evaluation → Conclusion

01. OBSERVE

Notice real behavior, friction, anomalies, and unspoken workarounds in everyday environments, interfaces, and human interactions before imposing categories or solutions.

02. IDENTIFY PATTERNS

Look across isolated events to see recurring structures—where similar failures, incentives, or behavioral responses appear across different domains or user groups.

03. FORMULATE THE QUESTION

Translate a vague observation into a precise, testable research inquiry. The quality of understanding is bounded by the clarity of the question asked.

04. INVESTIGATE

Examine the problem through multiple lenses—user behavior, interface mechanics, technical constraints, organizational incentives, and existing research.

05. DEVELOP A FRAMEWORK

Organize findings into a coherent conceptual model that explains how the variables relate, making complex dynamics legible and actionable.

06. EVALUATE

Test the framework against evidence, edge cases, counterexamples, and competing explanations to see where the model holds and where it breaks down.

07. CONCLUDE PROVISIONALLY

Arrive at a clear working position that can guide product decisions or further writing, while remaining open to revision when better evidence emerges.

CORE PRINCIPLE

I want to understand the problem behind the screen before deciding what the screen should look like.

LAYER 01

SURFACE

Visual hierarchy, typography, layout, color, components, information presentation.

LAYER 02

INTERACTION

Navigation, actions and responses, interaction states, feedback, error prevention, recovery.

LAYER 03

SYSTEM

Information architecture, data relationships, permissions, business rules, workflows, integrations, dependencies.

LAYER 04

OUTCOME

Task completion, clarity, decision support, user and business value, unintended consequences.

03.1 RESEARCH CATEGORY

PRODUCT & MARKET RESEARCH

Product discovery
Problem definition
Market conditions
Competitor analysis
User needs and pain points
Business objectives
Product opportunities
Market positioning
Product strategy

Examines why a product should exist, what structural gap exists in the market, how alternatives solve or fail the problem, and how user value aligns with business strategy.

03.2 RESEARCH CATEGORY

UX RESEARCH & USER UNDERSTANDING

User needs and expectations
User journeys
Task analysis
Usability
Cognitive load
Accessibility
User feedback
Interaction behavior

Focuses on how real people perceive tasks, where mental models diverge from system logic, how much cognitive effort an interface demands, and what barriers prevent completion.

03.3 RESEARCH CATEGORY

SCREEN RESEARCH & INTERFACE ANALYSIS

Visual hierarchy
Information architecture
Interaction patterns
Layout and composition
Information density
System feedback
Responsive behavior
Error states
Design systems

Deconstructs screens and component systems to evaluate how information is prioritized, how states are communicated, and how edge cases or errors are handled.

03.4 RESEARCH CATEGORY

CLIENT, CUSTOMER & STAKEHOLDER RESEARCH

Business requirements
Customer expectations
Stakeholder priorities
Operational workflows
Organizational constraints
Conflicting requirements
Product success criteria

Investigates the operational reality behind the product—how teams work, where stakeholder incentives conflict, what constraints exist, and how success is defined.

03.5 RESEARCH CATEGORY

BEHAVIORAL & AUDIENCE RESEARCH

Audience behavior
Content performance
Attention and engagement
Visual communication
Content distribution
Audience preferences
Creative experimentation

Studies how people allocate attention in information-dense environments, what earns trust or sustained engagement, and how framing and distribution shape response.

CENTRAL TENSION · OPEN INQUIRY

More capability does not automatically mean better human control.

DOSSIER 04.1

HUMAN AGENCY

How do we preserve genuine human agency when intelligent systems shape options, defaults, and recommendations?

Investigates the difference between nominal choice and informed self-direction when software pre-filters possibilities or frames the decision space.

Exploration: Choice architecture, default bias, intent preservation, reversibility.

Related: 05 Working Theories 12 Writing

DOSSIER 04.2

MEANINGFUL OVERSIGHT

What makes human-in-the-loop oversight genuinely meaningful rather than a rubber-stamp ritual?

Examines why placing a human reviewer at the end of an automated pipeline fails when the reviewer lacks context, time, or epistemic visibility.

Exploration: Automation complacency, verification friction, confidence calibration.

Related: 05 Working Theories 15 Practice

DOSSIER 04.3

AGENTIC AI

How should interaction paradigms change when software transitions from responding to prompts to executing multi-step actions?

Explores how boundaries, permissions, checkpoints, and observability must be designed when autonomous agents act across workflows.

Exploration: Scoped autonomy, action previews, interruptibility, rollback mechanisms.

Related: 10 Future of Building 12 Software Acting

DOSSIER 04.4

COGNITIVE DELEGATION

What happens to human judgment, skill retention, and critical thinking when we delegate synthesis and reasoning to AI?

Studies the trade-off between immediate cognitive relief and long-term human capability when people offload reading, framing, and evaluation.

Exploration: Skill atrophy vs. augmentation, epistemic vigilance, productive friction.

Related: 07 Psychology 12 Research Notes

DOSSIER 04.5

AI DECISION-MAKING

How should probabilistic outputs and uncertainty be communicated when AI informs consequential human decisions?

Investigates how fluent language or clean scores can mask model uncertainty, missing data, or brittle assumptions.

Exploration: Uncertainty visualization, reasoning transparency, provenance.

Related: 02 Product Layers 13 Idea Evaluation

DOSSIER 04.6

AI GOVERNANCE & SAFETY

How do product-level design decisions connect to institutional accountability, safety boundaries, and societal impact?

Connects high-level safety principles to concrete product architecture—where permissions, escalation paths, and liability actually live.

Exploration: Responsibility attribution, organizational guardrails, failure containment.

Related: 06 Systems Thinking 09 Geopolitics

WORKING THEORIES — NOT ESTABLISHED CONCLUSIONS

THEORY 01

1. Formal control is not meaningful control.

Giving a user a Confirm or Override button satisfies formal control on paper, but if the user lacks the time, context, or ability to evaluate the underlying system state, that control is purely symbolic.

Question Investigated: What interface conditions are required for human control to be practical rather than nominal?

THEORY 02

2. Approval is not the same as oversight.

Reviewing a polished final output generated by an automated system often invites passive agreement. True oversight requires visibility into the assumptions, excluded alternatives, and confidence boundaries that shaped the output.

Question Investigated: How does the design of review workflows change whether humans detect subtle errors or rubber-stamp them?

THEORY 03

3. Delegation changes responsibility.

When tasks are delegated to software or AI agents, people's sense of authorship and moral responsibility shifts—even when organizational rules state that the human remains fully accountable.

Question Investigated: How does delegating cognitive and operational work alter perceived ownership of outcomes?

THEORY 04

4. More capability requires better boundaries.

As systems become more capable and autonomous, the primary design challenge shifts from making actions easier to defining clear boundaries, permissions, escalation thresholds, and recovery paths.

Question Investigated: How should product architecture constrain and structure high-capability systems so they remain trustworthy?

LAYERED SYSTEMS VISUALIZATION (RECIPROCAL INTERACTION)

SYSTEMS DIAGNOSTIC QUESTIONS

1. Where does the problem originate?

2. Which people or systems are affected?

3. What dependencies influence the outcome?

4. Which incentives shape behavior?

5. What happens when one part changes?

6. Could a proposed solution create another problem elsewhere?

01. Perception and attention +

Human attention is selective and finite. What people notice first, what they filter out, and how visual or informational salience shapes interpretation determines how any interface or message is actually experienced.

02. Cognitive biases +

Mental shortcuts—such as anchoring, confirmation bias, status-quo preference, and framing effects—help people navigate complexity quickly, but also introduce predictable blind spots in reasoning and product interaction.

03. Decision-making +

Real-world decisions rarely happen with complete information or unlimited time. Investigating how people weigh uncertainty, risk, effort, and regret helps explain why actual behavior often diverges from rational models.

04. Social influence +

People do not decide in a vacuum. Social proof, authority cues, group norms, status signaling, and institutional trust strongly influence what individuals accept as credible, normal, or desirable.

05. Empathy and moral attention +

Our capacity to care about others is real, yet unevenly distributed. Understanding how proximity, identity, narrative framing, and visibility direct moral concern is essential for understanding both human relationships and collective behavior.

06. Human–technology relationships +

As software becomes conversational, adaptive, and autonomous, people project social expectations, trust, or authority onto machines—reshaping how humans rely on technology and how technology shapes human habits.

INDEPENDENT MONOGRAPH · IN PROGRESS (UNPUBLISHED)

SELECTIVE EMPATHY — Why does our concern for people vary?

Selective Empathy is an independent writing and research project examining why human compassion, moral urgency, and outrage are strongly activated in some situations while remaining muted in others. Rather than treating empathy only as an individual trait, it investigates the structural factors—proximity, similarity, visibility, group identity, media framing, and collective institutions—that govern how moral attention is distributed.

CENTRAL OPEN INQUIRY

Can a common architecture help explain how people distribute concern, loyalty, trust, justice, forgiveness, and moral judgment?

TEN CORE THEMES

01. Proximity and psychological distance

02. Similarity and social identity

03. Visibility and attention

04. Empathy and moral judgment

05. Loyalty and group belonging

06. Trust and forgiveness

07. Tribalism and nationalism

08. Religion and collective identity

09. Media and polarization

10. Markets, war, democracy, and international relations

09.1 SUBCATEGORY

BUSINESS & STRATEGY

Business models
Value creation
Market dynamics
Competitive positioning
Product strategy
Organizational incentives
Industry transformation

09.2 SUBCATEGORY

ECONOMICS

Economic systems
Market structures
Incentives and behavior
Resource allocation
Technological change
Economic forces and business decisions

09.3 SUBCATEGORY

GEOPOLITICS

International relations
Global institutions
Technology and national power
Trade and industrial policy
Resource dependencies
Global competition
Politics, economics, and technology

My engagement with AI operates on two levels simultaneously: using AI as a practical capability to accelerate discovery, ideation, and 0→1 prototyping, while researching AI as a design problem requiring legibility, human direction, and accountability.

01. AI-assisted product discovery

Synthesizing complex research inputs while preserving human nuance.

02. AI-assisted ideation

Expanding exploration across divergent structural possibilities.

03. Prototyping and product building

Moving rapidly from conceptual architecture to functional 0→1 prototypes.

04. AI capabilities and limitations

Distinguishing probabilistic generation from deterministic product requirements.

05. Understandable automated actions

Making system reasoning, actions, and state changes legible to users.

06. Human direction

Designing clear steering, scoping, and override mechanisms.

07. Accountability in automated workflows

Ensuring auditability, error recovery, and clear human responsibility across automated pipelines.

STAGE 01 — OBSERVE

Gather unfiltered signals from real user behavior, existing workflows, domain constraints, and system friction before jumping to solutions.

1. AI & Human Agency+

Central Question: How do intelligent defaults, predictive ranking, and automated suggestions reshape human autonomy without explicitly removing choice?

2. What Actually Makes a Product?+

Central Question: Beyond visual polish and isolated features, what structural relationships between problem, workflow, system logic, and outcome define a real product?

3. Cognitive Delegation+

Central Question: What happens to human comprehension and evaluative capability when we delegate synthesis, framing, and reasoning to machines?

4. The Interface Is Not the System+

Central Question: How do hidden data models, permissions, business rules, and operational dependencies determine what an interface can honestly promise?

5. Selective Empathy+

Central Question: Can a common architecture help explain how people distribute concern, loyalty, trust, justice, forgiveness, and moral judgment?

6. When Software Starts Acting+

Central Question: How must interaction design, permissions, and accountability evolve when software moves from executing commands to pursuing goals autonomously?

NOTES FROM THE DESK — SHORT OBSERVATIONS & DEVELOPING IDEAS

DESK NOTE 01 / FRICTION

Not all friction is a usability flaw. Sometimes a deliberate pause is the only thing keeping a person aware of the consequence of an action.

DESK NOTE 02 / CERTAINTY

When an interface displays a probabilistic guess with the same visual confidence as a verified fact, the interface is making an epistemic choice for the user.

DESK NOTE 03 / SIMPLICITY

Simplifying a screen often moves complexity elsewhere. The real question is whether complexity was resolved or merely hidden.

My goal is to remain curious about an idea without becoming attached to it.

01. Observation → What actually happened, independent of my assumptions?+

Separating raw phenomena from immediate narrative interpretation.

02. Interpretation → How else could this same pattern be explained?+

Surfacing alternative explanations before locking into a single story.

03. Hypothesis → If this idea is true, what specific outcomes must follow?+

Stating the claim clearly enough that it can be tested and falsified.

04. Evidence → What concrete evidence supports this, and how reliable is it?+

Distinguishing strong behavioral evidence from anecdote or intuition.

05. Counterexample → What evidence would force me to revise or abandon this?+

Actively seeking disconfirming cases where the model breaks down.

06. Limitations → Where are the boundary conditions of this idea?+

Defining the contexts where the explanation applies and where it stops.

07. Conclusion → What is the most accurate working conclusion right now?+

Committing in proportion to the evidence while remaining open to revision.

UNIFYING THEME

Understanding the relationships between people, decisions, and the systems around them.

Every domain in this archive investigates the same core dynamic at a different zoom level: how people perceive, decide, and act inside structured environments.

15.1 PRACTICE

PROBLEM FRAMING

Understanding the structural cause of a problem before committing to interface design.

15.2 PRACTICE

INFORMATION ARCHITECTURE

Organizing complex systems, hierarchies, and permissions so they align with human mental models.

15.3 PRACTICE

INTERACTION DESIGN

Designing clear states, feedback, error prevention, and recovery flows.

15.4 PRACTICE

AI PRODUCT DESIGN

Structuring human–AI workflows around agency, uncertainty legibility, and meaningful oversight.

15.5 PRACTICE

PRODUCT STRATEGY

Connecting 0→1 product decisions to market dynamics, incentives, and business viability.

15.6 PRACTICE

DESIGN EVALUATION

Evaluating outcomes by clarity, decision support, and systemic impact rather than surface aesthetics alone.

01.How do we design products that expand human capability without quietly eroding human judgment and agency?
02.What distinguishes meaningful human oversight of intelligent systems from symbolic approval?
03.When software shifts from responding to commands to taking multi-step actions, how should boundaries, permissions, and reversibility be structured?
04.How much cognitive work can people delegate to AI before losing the ability to evaluate whether the AI's output is sound?
05.How can interfaces represent uncertainty, missing context, and model limitations without overwhelming the person making a decision?
06.Why do well-intentioned product and organizational metrics frequently produce unintended behavioral consequences?
07.Can a common conceptual architecture explain why human empathy, loyalty, trust, and moral concern are distributed so unevenly?
08.How do digital media environments and algorithmic curation alter psychological distance, polarization, and collective identity?
09.How will economic incentives, industrial policy, and geopolitical competition shape the next decade of AI and software architecture?
10.As AI lowers the cost of generating interfaces and code, what remains the irreducible core of product thinking and design judgment?

01. Understand before solving.+

Resist jumping to interface solutions before understanding the origin and structure of the problem.

02. Look beyond the interface.+

Treat every screen as one layer connected to workflows, rules, incentives, and human consequences.

03. Question assumptions.+

Examine inherited requirements and default conventions to see whether they genuinely hold.

04. Connect disciplines.+

Draw across design, psychology, systems thinking, technology, economics, and society.

05. Separate evidence from interpretation.+

Distinguish what has been observed from what is inferred or hypothesized.

06. Preserve human agency.+

Design technology that supports human judgment, legibility, and clear responsibility.

07. Treat conclusions as revisable.+

Hold working theories with rigor while remaining ready to revise them when better evidence appears.

08. Turn understanding into something useful.+

Connect research and systems thinking back to building clear, resilient products.

A little about me.

I turn complexity into clarity.

I build products from 0 to 1.

I explore human–AI interaction.

I build beyond the screen.

I research people, behaviour, and systems.

EDUCATION & PROFESSIONAL DEVELOPMENT

Bachelor of Design (B.Des)

McKinsey.org Forward Program

IoT Product Management

AI – Data Engineering

Product Manager

Enterprise Governance

Design with Inclusivity

PROFESSIONAL EXPERIENCE

Polymer Bazaar

NetCom Learning

CRMS Tech

1. Formal control is not meaningful control.

2. Approval is not the same as oversight.

3. Delegation changes responsibility.

4. More capability requires better boundaries.