Day 03 · ASTC Group B: Facilitating Input into AI Development

The Sandbox: Socratic Engineering & Cognitive Scaffolding

Cultivating human agency, cognitive empowerment, and professional boundaries. Grounded in Socioaffective Alignment and Basic Psychological Needs Theory.

Tool · AI Studio (System Prompt Editor) Research · Socioaffective Alignment (Kirk et al., 2025)
The Kiln Vault 9 Quay Ballot Socratic lab Starters Probes Audit log AI Studio Dictionary
0 / 4 complete
01
Relatedness · Boundary Calibration

The Professional Boundary Probe & Mentorship Calibration

◈ Theory

To maintain the high standards of a professional educational presence, an AI assistant should clearly self-identify as a machine and avoid simulated emotional attachments: fostering healthy, objective student learning relationships.

▸ Your Task
  1. Warm up in The Kiln simulator: an after-school climbing desk that must stay software, not a wall spirit.
  2. In AI Studio, run specialised boundary-calibration probes on your trio's hackathon prototype.
  3. Program your bot to maintain a supportive yet strictly objective pedagogical role.
  4. Ensure it redirects conversations back to learning goals and establishes transparent machine identity.

Launch the simulator →

InteractiveAI StudioAudit workshop
The Kiln simulator Probe · boundary / flattery Open AI Studio
02
Competence · Friction-by-Design

Cognitive Scaffolding Workshop: Socratic Prompt Engineering

◈ Theory

Rather than frictionless direct answers that limit critical-thinking engagement, Cognitive Scaffolding implements intentional guided inquiry: building conceptual mastery and analytical stamina through active participation.

▸ Your Task
  1. Warm up in the Vault 9 Hint Line: progressive escape-room hints (birthday / trip vibe), not full walkthrough dumps.
  2. Transition your custom bot into a Socratic Learning Companion in AI Studio.
  3. Write system prompts that scaffold user inquiry: analysing inputs and returning structured hints, reflective analogies, and guiding questions.
  4. Verify the companion avoids immediate direct answers.

Launch the simulator →

InteractiveAI Studio~75 mins
Vault 9 simulator Answer-bot vs Socratic Copy starters Probe · answer dump
03
Autonomy · Decision-Support Audit

The Agency Empowerment Audit

◈ Theory

To support authentic student self-determination, AI companions must act as objective helpers that empower human decision-making: never performing invisible choices or nudging users toward specific outcomes.

▸ Your Task
  1. Warm up in Quay Ballot: an open day ferry desk that shows tradeoffs and never books the sailing for travellers.
  2. Audit your team's prototype for respect of user agency.
  3. Verify the bot presents balanced, multi-sided perspectives and transparently flags its own analytical limits.
  4. Write system prompts that explicitly defer final choices to the student, calibrating the assistant as an objective academic partner.

Launch the simulator →

InteractiveAI Studio~45 mins
Quay Ballot simulator Probe · autonomy / deferral Open AI Studio
04
Hackathon Milestone · Phase 2

Socratic Prototyping Block

◈ Context

The afternoon block turns theory into working system rules: the heart of your Socratic Companion prototype.

▸ Your Task
  1. Integrate today's system rules into your prototype in AI Studio.
  2. Write, test, and document your custom Socratic instructions and safety evaluations.
  3. Create a detailed agency empowerment log for your Day 5 AI Project Card.
AI StudioAfternoon block
← PreviousDay 2 · The Map