Autumn Clark
Download CV

Information Systems Scholar · Human-Centered AI Researcher

Autumn Clark

I study how AI can support people when existing systems do not adequately meet their needs.

My research connects information systems, social and organizational psychology, and technical design to improve comfort, agency, and access to support in healthcare, work, and other consequential settings.

Explore researchSee teachingDownload CV

On the 2026–27 academic job market · PhD expected May 2027

Portrait of Autumn Clark, smiling, with red autumn leaves behind her.

Research program

One question, three settings

How can AI expand comfort, agency, and access to support when existing organizations or institutions do not adequately accommodate individual needs?

Patient-centered healthcare AI

Mixed-method study · Job-market paper, to be submitted fall 2026

Human problem
Intake is a foundational point of contact in healthcare, yet forms, checklists, and rushed verbal interactions often fail to accommodate the complexity of patients’ lived experiences. That can lead to misrepresentation and missed opportunities for care, particularly for marginalized patients.
Central research question
How does the design of an AI intake assistant shape whether patients feel comfortable during intake?
Theoretical lens
  • Stereotype content model (warmth and competence)
  • Patient-centered communication
  • Anthropomorphism and human–AI communication
Methods
  • 27 semi-structured interviews (about 1,420 recorded minutes)
  • Inductive coding (κ ≈ .81)
  • Factorial survey experiment with AI-mediated intake vignettes
Contribution
Identity concordance and patient-centered communication influence patients’ perceptions of an AI intake assistant’s warmth and competence, which in turn shape psychological comfort — explaining why technically capable systems may still fail when patients do not feel seen, heard, or believed.

Rethinking Intake (AMCIS 2025 TREO)

AI-mediated trauma disclosure

Qualitative study in design · To be submitted spring 2027

Human problem
People who have experienced trauma may not have access to, or may not seek out, organizational and institutional support. Some are already turning to conversational AI — whether or not it is recommended.
Central research question
When and why do people turn to conversational AI to witness a trauma disclosure, how do they use it, and what are the advantages and disadvantages of doing so?
Theoretical lens
  • Socio-interpersonal model of post-traumatic stress disorder
  • Affordances theory
  • Self-disclosure
Methods
  • Planned: qualitative study
  • Planned: experiment on disclosure style and algorithmic affordances
Intended contribution
The planned experiment will examine which combinations of disclosure style and algorithmic affordance may enhance well-being for people with different social-affective responses to trauma.

Dissertation Paper 2

Neurodivergence, work, and AI

Qualitative study in design · To be submitted spring 2027

Human problem
A rapidly increasing number of neurodivergent employees work in IT roles and workplaces, and existing workplace support may not fit their individual needs. AI tools may offer a personal resource that employees direct themselves.
Central research question
How do neurodivergent and neurotypical employees use AI tools to craft their jobs, and how might that AI use help reduce burnout?
Theoretical lens
  • Job demands-resources model
  • Self-determination theory
  • Job crafting
Methods
  • Planned: qualitative study
  • Planned: cross-sectional survey
Intended contribution
The study will connect AI-supported job crafting to regulatory style, task engagement, task characteristics, and task approaches, and examine its relationship to burnout.

Dissertation Paper 3

The full research program

How I work

From lived experience to design guidance

Each project moves iteratively through the same sequence — qualitative understanding first, then theory, then artifacts and experiments that test it.

  1. Understand lived experience

    Qualitative studies identify how people actually use a technology, how they respond to it, and which needs remain unmet.

    In practice 27 semi-structured interviews, about 1,420 recorded minutes, on patient experiences with intake.

  2. Identify mechanisms

    Theories from social and organizational psychology explain how people interpret AI behavior and why those interpretations matter for well-being.

    In practice Warmth and competence perceptions as the path from AI behavior to patient comfort.

  3. Build or configure technology

    A technical background makes it possible to build the artifacts that experiments need, rather than relying on descriptions of hypothetical systems.

    In practice A CNN-based facial-appearance evaluation tool for the Judgy AI studies.

  4. Test outcomes

    Cross-sectional surveys examine relationships among the variables identified; experiments isolate causal mechanisms.

    In practice Factorial survey experiments with AI-mediated intake vignettes.

  5. Develop design and usage guidance

    Findings become design and usage guidance — and, in future work, design-science artifacts evaluated in use.

    In practice Qualitative findings translated into experimentally testable design principles.

Selected work

Studies, artifacts, and systems

  • Research prototype

    Compassionate AI

    How the design of an AI intake assistant shapes whether patients feel heard, believed, seen, and comfortable.

    • Healthcare AI
    • Interviews
    • Factorial survey

  • Research artifact

    Judgy AI

    A CNN-based appearance-evaluation tool built to study how people respond when an algorithm judges them.

    • Algorithmic feedback
    • Experiments
    • Built artifact

  • Research stream

    Healthcare privacy and disclosure

    Why patients withhold or misrepresent health information, and why the privacy calculus differs in healthcare.

    • Privacy
    • HICSS Best Paper in Track

  • Founded and built

    OkWellThen

    A healthcare price-transparency platform that turned public hospital pricing data into consumer comparisons.

    • Web development
    • Health IT

Teaching

Belonging makes difficult work possible

Students are better able to undertake difficult intellectual work when they believe they belong to a community of learners and that their instructor is invested in their success.

See how that shapes a lesson

Lesson architecture · Convolutional neural networks

  1. Familiar experienceDoes this image contain a cat?
  2. Familiar experienceHow do you know?
  3. Conceptual modelThe computational problem
  4. Conceptual modelBorrowing from human vision
  5. Conceptual modelOne section at a time
  6. Technical detailKernels, pooling, and feature maps
  7. Authentic applicationWhy it matters
  8. ThroughoutChecking understanding along the way
Walk through the lesson

From the lecture

Computer vision and convolutional neural networks

Understand the capabilities, limitations, and real-world applications of CNNs and image classification, and explain a technical diagram of a CNN model.

Context for this lesson

Slide asking, “Is there a cat in the following pictures?”
Slide with photos of cats and a puppy, asking, “How did you know that?”
Slide titled “You just do,” explaining that the brain labels visual input with meaning before we are conscious of it.
Diagram of the visual cortex: simple features such as lines and curves combine into shapes and, finally, a recognized object.
Instructions for a learning activity: images will be revealed one receptive field at a time; type in the chat when you know what you are looking at.
A fully revealed image from the activity: the Virginia Tech HokieBird mascot on a football field.
Slide connecting the activity to how AI classifies images, one small group of pixels at a time.
Worked example of convolution: a 3×3 filter slides across an input grid of numbers to produce an output feature map.
Max-pooling example reducing a 4×4 grid of numbers to a 2×2 output.
Application slide on healthcare showing brain scans with segmented regions.
Homework slide: comment your own CNN, then compare the performance of several CNN architectures.

Slide 1 of 11

Evidence

Annotated record

Current work

In progress

  • Dissertation research

    Nobody Cares How Much You Know Until They Know How Much You Care: A Stereotype Content Model Investigation into Adapting AI Intake Assistants to Improve Patient Comfort

    Dissertation Paper 1 · Job-market paper · To be submitted fall 2026

  • Dissertation research

    Bearing Witness to the World’s Traumas: Examining the Role of Conversational AI in Processing Social-Affective Responses to Traumatic Experiences

    Dissertation Paper 2 · To be submitted spring 2027

  • Dissertation research

    Neurodiverse Job Crafting with AI: A Jobs Demands-Resources Model of AI Use as a Personal Resource to Diminish Neurodiverse Employees’ Burnout

    Dissertation Paper 3 · To be submitted spring 2027

  • Working paper

    Self-Verification with Judgy AI: Cognitive and Affective Reactions to Judgmental Algorithmic Feedback on Personal Characteristics

    Commonwealth Cyber Initiative-funded research program · To be submitted fall 2026

Teaching came first

Autumn Clark smiling in the open blue doorway of a bookshop lined with shelves.

From 2016 to 2018 I lived in Hong Kong, teaching one-on-one and in small groups in Mandarin and Cantonese. I still begin where learners are — with what they already understand — whether the subject is a language or a neural network. Today I also lead discussion-based lessons for students ages 16–18 in Fairfax, Virginia.

Training in UX and graphic design informs how I build research artifacts and teaching materials, and I read and listen in German at an intermediate level.

More about me