Assignments

The Ed board for posting your weekly writing is available here.

Due: Monday, March 9, at 5:00 PM

In our final class, we will cover what is arguably the most fundamental question: does AI pose an existential risk to humanity?

• Reading:

  • How to Stop Superhuman A.I. Before It Stops Us (Stuart Russell)
  • We Shouldn’t be Scared by ‘Superintelligent A.I.’ (Melanie Mitchell)
  • The Supposed Existential Threat of AI to Humanity (Oren Etzioni & Rich Chen)
  • OPTIONAL: How Afraid of the A.I. Apocalypse Should We Be? (Ezra Klein interviewing Eliezer Yudkowsky)
  • And a recommendation from your TA: How to know if artificial intelligence is about to destroy civilization (Oren Etzioni)

• Writing:

  1. Is AI an existential risk to humanity? If so, what should we do about it? (1–2 paragraphs)

Due: Monday, March 2, at 5:00 PM

• Reading:

  • How AI Destroys Institutions. Note: don't be intimidated by the length of the document. It's easy reading, and you can ignore the extensive footnotes which often take up much of the page.
  • Also, an article on Anthropic's pentagon involvement. You should be able to log in with your UW email, but in case you can't, I've attached a google docs version here.

• Writing:

  1. Find one risk you agree with and offer your own observation as to why (1 paragraph)
  2. Find another risk you disagree with and explain why (1 paragraph)

Due: Monday, February 23, at 5:00 PM

• Reading:

  • Dario Amodei's essay: "The Adolescence of Technology"

• Writing:

  1. What is Dario saying about slowing AI progress? Agree or disagree? (1 paragraph)

Due: Monday, February 16, at 5:00 PM

• Reading:

  • Dario Amodei's essay: "The Adolescence of Technology"

• Writing:

  1. How would you address the autonomy risk articulated by Dario? (1 paragraph)

Due: Monday, February 9, at 5:00 PM

• Reading:

  • 8 perspectives on the future of AI captured in one-sentence sound bites: NYT Opinion survey

• Writing:

  1. Disagree with one of the points made using an original argument. (1 paragraph)

Due: Monday, February 2, at 5:00 PM

• Reading:

  • Newell & Simon's Turing Award Lecture: "Computer Science as Empirical Inquiry: Symbols and Search"
  • Wing on Computational Thinking (choose one):
    • 2006 original CACM article
    • 2010 expanded version

• Writing:

  1. What is a fundamental law of qualitative structure for AI? (1 paragraph)
  2. What would be instances of "AI thinking"? (1 paragraph)

Due: Monday, January 26, at 5:00 PM

• Reading:

  • Banko article: paper (skim to understand Figure 1)
  • Sutton's "bitter lesson": essay
  • Brooks' response - the "better lesson": essay
  • OPTIONAL: MIT Technology Review article

• Writing:

  1. Who do you agree with, Brooks or Sutton? Why? (1 paragraph)

Due: Monday, January 19, at 5:00 PM

• Reading:

  • Melanie Mitchell's 2019 book, Artificial Intelligence: A Guide for Thinking Humans, was named one of the five best books on AI by both The New York Times and The Wall Street Journal in 2024. Chapter 1
  • Jeff Dean (UW PhD, head of Google AI) recounts the familiar history of Deep Learning with some "twists".
    "A Golden Decade of Deep Learning: Computing Systems & Applications" by Jeffrey A. Dean, 2022 (17 pages)

• Writing:

  1. What resulted in previous AI Winters? Will we see another one in the next 3-5 years? Why (or why not)? (1 paragraph)
  2. What is a surprising/novel point that Jeff Dean makes? (1 paragraph)

Due: Tuesday, January 13 - before class

• Reading:

  • Brief overview: Turing Test and Chinese Room
  • Turing Test original: paper (long, feel free to skim)
  • Recent assessment: Nature article - log in with your UW email to access.
  • OPTIONAL: On the limitations of benchmarks: paper

• Writing:

  1. Do you believe in strong or weak AI? Give a crisp (single) reason for your belief. (1-2 paragraphs)
  2. What version of the Turing Test (if any) is useful for assessing AI systems through 2030? (1-2 paragraphs)