🏛️ Issues and Advocacy Training
Courses

Artificial intelligence

⚙️ AI Issues: Technical, Legal, Social and Logistical

Understand artificial intelligence with your feet on the ground: what it runs on and costs, what it uses in power and water, the copyright fights, bias, privacy, jobs, safety debates and the laws as of October 2026, plus how to test an AI claim before you repeat it or act on it.

9 lessons · about 2.5 hours ·

Start lesson 1

  1. 1What AI Runs On: Chips, Compute and CostFollow an AI answer back to the chips, buildings and money behind it, and learn to tell training costs from the cost of everyday use.16 min
  2. 2Energy and Water: What Is Measured and What Is GuessedSort AI energy and water numbers into measured, company-reported and modeled, and learn why per-query estimates range over a factor of a hundred.17 min
  3. 3Training Data and Copyright: The Cases So FarLearn where training data comes from, how US fair use works, and where the major AI copyright cases stood as of October 2026.18 min
  4. 4Bias and Errors: How AI Gets Things WrongTrace how bias gets into AI systems, look at documented cases, and learn what hallucination rates measure and what they leave out.16 min
  5. 5Privacy and Security: Your Prompts, Injection and DeepfakesFind out what happens to what you type into an AI tool, how attackers hijack AI through hidden instructions, and how deepfake fraud works.16 min
  6. 6Jobs and the Economy: What the Research Shows So FarSeparate measured effects of AI on work from forecasts, and learn why exposure, productivity and job loss are three different numbers.16 min
  7. 7AI Safety and Risk: Who Argues WhatMap the main positions in the AI risk debate, from present-day harms to catastrophic risk, and learn what evidence each side relies on.16 min
  8. 8AI Law and Regulation as of October 2026Get the current state of AI rules in the EU, at the US federal level and in the states, including Idaho's election deepfake law and Colorado's rewritten AI Act.18 min
  9. 9Checking an AI Claim: Benchmarks, Demos, Vendors and RecordsTest AI claims the way you would test any other: read benchmarks skeptically, separate a demo from daily use, ask vendors and agencies the right questions, and request the records.18 min