About this site
Purpose, the exam and certificate, privacy, disclaimer, and sources.
Purpose
A learning site for understanding the large-scale risks AI may bring, without either fear-mongering or complacency, so that people can prepare. We do not cover attack methods; the focus is prevention, preparedness and response.
The exam and certificate
A digital certificate (PDF) is issued for 85 or more correct answers out of 100. It shows completion of learning and is not a public qualification or license. Scoring takes place inside your browser.
Privacy
Your name, answers and checklist are processed and stored only in the browser on your device and are never sent to a server. The certificate PDF is also generated on your device.
Disclaimer
Scenarios and time frames are forecasts for learning, not a fixed future. Probability figures are individual experts' subjective estimates. Information is as of September 2026. In an emergency, follow official information from your government and local authorities.
Main sources
- International AI Safety Report 2026
- Axios (Sept 9, 2026): the AI extinction debate and experts' probability estimates
- Fortune (July 21, 2026): disclosure that AI models under test escaped their isolated environment
- Jensen Huang's "0% chance by 2030" remark (Sept 2026)
- AI Futures Project: updated AI timeline forecasts (Jan 2026)
- Jehn et al., "No Place to Hide?" — commentary on the research review (2026)
- Boyd & Wilson: island refuges for surviving nuclear winter (Risk Analysis)
- Boyd & Wilson: island nations as refuges from extreme pandemics
- CAIS "Statement on AI Risk" (2023); Japan's Ministry of Agriculture (food self-sufficiency) and Agency for Natural Resources and Energy (energy self-sufficiency)