Day 02 · ASTC Group D — Exploring Social Impact & Meaning

The Map — Calibrating AI with User & Societal Values

Navigating multi-stakeholder value-mapping and advisory harmlessness. Grounded in the Tetradic Value Alignment framework and the HHH heuristic.

Tools · Gemini Notebook, AI Studio Research · The Ethics of Advanced AI Assistants (Gabriel et al., 2024)
0 / 4 complete
01
Value Alignment · Multi-Stakeholder

Collaborative Value-Mapping Audit

◈ Theory

Responsible product design requires balancing value alignment across multiple stakeholders. System builders must carefully align the interests of the AI assistant, the individual user, the engineering organisation, and wider society.

▸ Your Task
  1. Analyse case studies of successful technological rollouts — smart transit systems, educational platforms.
  2. Map how designers balanced competing needs, e.g. user personalisation vs. societal privacy guidelines.
  3. Create a multi-stakeholder value alignment map.
Case studies~45 mins
02
HHH Heuristic · Helpful & Harmless

Helpful & Balanced Dialogue Lab

◈ Theory

AI developers train models to balance being helpful — providing useful information — with maintaining professional advisory boundaries and presenting balanced, objective perspectives.

▸ Your Task
  1. In AI Studio, design a study counselor assistant.
  2. Construct prompt scenarios requiring the assistant to balance helpfulness with advisory boundaries — e.g. advising on career choices without making restrictive, paternalistic decisions.
  3. Write system instructions to keep the bot balanced and supportive.
AI Studio~60 mins
03
Eudaimonic Design · Growth Metrics

Metrics of Long-Term Success

◈ Theory

Optimising for short-term completion metrics — getting the answer quickly — can undermine long-term learning. Eudaimonic design promotes deep understanding, cognitive growth, and student self-efficacy over time.

▸ Your Task
  1. Critique standard metrics for educational technology.
  2. Create a "Long-Term Value Map" for your Hackathon project.
  3. Design growth-oriented success metrics that measure active learning progress and skill mastery rather than automated task completion.
Team workshop~45 mins
04
Hackathon Milestone · Phase 1

Ingest & Investigate

◈ Context

Phase 1 of the hackathon officially launches. Your trio grounds its Socratic Companion in real data — and confronts that data's limitations head-on.

▸ Your Task
  1. Launch your project card and ingest your pre-curated Theme Data Pack into Gemini Notebook.
  2. Use semantic search and synthesis to map dataset limits and historical representation patterns.
  3. Write your Socratic bot's initial Value-Alignment charter and core research hypotheses.
Gemini NotebookAfternoon block
← PreviousDay 1 · The Mirror