About the journal

Human Side

Human Side is Dr. Sarah Dyson’s journal on ethical AI leadership. The field notes stay with the people who still have to answer once an intelligent system is in the room.

Dr. Sarah Dyson is a psychologist and leadership scholar who works where intelligent systems meet the people who still have to answer for them. She is Program Lead and Assistant Professor in the Doctorate of Digital Enterprise Leadership at Harrisburg University of Science and Technology, and Associate Professor at National University.

She has taught at Harrisburg since 2020: graduate capstone, research methods, emotional intelligence for project managers, and requirements analysis. Before that she spent more than a decade as an entrepreneur and chief executive in mental health, and as a research consultant in public and private organizations.

Her research follows how AI, media, emotion, and motivation change the skills, attitudes, and behaviors of project managers and leaders. The record is on the Research page. She holds a Ph.D. in Psychology, with a specialization in social psychology, from Walden University, and an M.S. in Business Management and Information Management from Colorado Technical University. She is a member of the National Coalition of Independent Scholars.

Contact

For the journal, speaking, and questions about the work.

Based in
Jonesboro, Arkansas

Themes

These are the questions the journal returns to. New field notes move across them so the work does not collapse into a single subject.

  • Status / SCARF under AI

    How status, certainty, autonomy, relatedness, and fairness move when a team works beside a fluent agent.

  • Attachment and tool loss

    What happens to identity and relatedness when a familiar agent is retired, swapped, or silently degraded.

  • Trust as capital

    Treat trust as deposits, withdrawals, and calls — measurable, dated, and owned after a system recommends.

  • Explainability as a leadership skill

    What a non-technical leader actually needs to stand behind an opaque path — translation, not transparency theater.

  • Fairness in the loop

    Who gets seen when a shortlist, score, or routing path is touched by a model — quiet bias in fluent output.

  • Human overhead of agents

    The review load, exception handling, and unpaid vigilance that never appears in the vendor deck.

  • Named override

    An agentic path that can proceed without asking is unclaimed authority until a person, role, and channel are named.

  • Governance that fits in a week

    Controls people actually use this week, not a policy PDF. Measures that detect trust withdrawal and agency loss.

  • Rituals after the model moves

    Handoffs, reviews, and stand-ups still required when automation runs — keeping a person on the hook.

  • Near-miss and rollback

    Pilot failure, scale-back, rehire, or the miss that never entered the log — what the metrics missed.

  • ADHD / neurodivergent scaffolding

    AI as executive-function support in project management and leadership — strengths-based use, evaluation bias, load.

  • EI under AI load

    Self-awareness, regulation, and empathy when tools accelerate work — priorities leaders still under-build while buying software.

  • Retiring an agent without retiring the team

    How to take a system out of production without taking the team's competence with it.

For the portable pause before a system starts to sound like a colleague, see the Toolkit.