Skip to main content

Sources and attribution

Reading map and attribution

Core readings and offline equivalents

Module 1: International AI Safety Report summary, introduction and risk categories; alternate: module 1's three claim cards.
Module 2: METR limitations note; alternate: W2's evidence cards and evaluation design.
Module 3: International AI Safety Report summary, risk sections; alternate: module 3's classification cases and W3.
Module 4: International AI Safety Report summary, risk-management section; alternate: W4's safeguard cards.
Module 5: BlueDot contribution ideas; alternate: module 5's three project options.

Readings support 30-minute source labs, not an additional full course. If a link is unavailable, use the named in-package alternative. The official report summary is available in English and Spanish. Other external readings are primarily English; HASI's equivalent core tasks and explanations are fully bilingual.

Source links

BlueDot AGI Strategy curriculum
BlueDot independent reuse guidance, Joshua Landes, 2026
BlueDot shared AGI discussion template
BlueDot activity templates
BlueDot facilitator training
International AI Safety Report 2026: executive summary
METR: Clarifying limitations of time horizon, January 2026
Narayanan and Kapoor: AI as Normal Technology, April 2025

What HASI adapted

Retained: the five-stage progression, prepared discussion, peer explanation, causal risk analysis, layered defenses, and a personal contribution plan. Added: original Puente cases, explicit distinction between fact and scenario, college-level scaffolding, bilingual alternatives, assessment anchors, and host procedures. The shared BlueDot scenarios and their numerical future outcomes are not presented as forecasts or reused as factual evidence.

BlueDot's materials are supplied without warranty under its stated reuse conditions. Original third-party articles are linked, not reproduced or fully translated. The course hub may require a free account for some content; HASI's core activities do not depend on that account. Participant edition 1.1, September 20, 2026; sources checked September 12, 2026.