Assignments
📝 HW1 – Due: Tue, Jan 13, 1:59 PM PT (before class)
HW1¶
Task
Read Alan Turing’s “Computing Machinery and Intelligence”.
Required: Sections 1–5
Optional: Remaining sections
Write a 200–350 word summary in your own words.
Note: We will discuss this paper in class on Tue, Jan 13.
Accessing the paper
Paper link: TURING (1950)
To access the paper:
Go to the link above and scroll to the Sign in section.
Click Sign in through your institution.
Under Find Your Organization, enter or select
University of California – San Diego (Library, OpenAthens).Log in using your UCSD account.
If you are unable to access the paper this way, a copy is also available on Canvas.
Optional guiding questions
You may find the following questions helpful as you write your summary. You do not need to answer them explicitly.
What question is Turing trying to address?
What are the main ideas or arguments he introduces?
How does Turing propose evaluating whether a machine can think?
What conclusions or implications does Turing suggest?
Grading
Graded for completion
Submission
Upload a PDF of your summary to Gradescope under
HW1
Late policy
Up to one week late accepted
Note on GenAI tools
For reading-based assignments like HW1, I want to be clear about expectations and about the thoughtful use of GenAI tools (e.g., ChatGPT, Gemini, etc.)
GenAI tools can easily generate a generic summary of the Turing paper. To be transparent and to avoid unnecessary duplication of effort, I’ve shared an example of such a summary below labeled GenAI Summary (Spoiler Alert).
If you are in a time crunch this week or are unable to engage deeply with the reading, you are welcome to copy that summary directly for the assignment. You will not be penalized.
(Another reason for sharing the summary here is that generating AI outputs requires nontrivial computational and energy resources; we’ll talk more about this later in the course.)
That said, my expectation is that you will go beyond the summary and engage directly with the paper.
For assignments like these, the point is not the written artifact itself. The point is that you read the paper, wrestle with the ideas, and form your own understanding. The learning comes from that process, not from the final text you submit.
Reading the summary first is completely fine. For many people (myself included), it can be a helpful way to ease into a dense paper. What matters is that your engagement does not stop there.
If you choose to use GenAI tools, the sequence matters:
You may start with a summary if that helps you orient yourself.
Then read or skim the paper itself at least once.
Form your own perspective, questions, points of agreement or disagreement, and areas of confusion.
Use GenAI tools afterward to clarify ideas, test your understanding, or explore alternative interpretations.
More productive uses of these tools include:
Asking questions about parts of the paper you found confusing.
Asking what Turing might mean by a particular claim and whether there are multiple interpretations.
Asking where in the paper a specific argument is made, and then going back to read that section yourself.
Asking what assumptions Turing is making or what someone might reasonably disagree with.
Asking for related papers or critiques, and then reading those sources.
Summaries are fine as a starting point. Engagement with the paper is what matters. My recommendation and expectation from you remains the same: read the paper, think about it, and write in your own words. That is where the value lies.
GenAI Summary (SPOILER ALERT!)
Here is a 200–350 word summary you can submit directly:
In “Computing Machinery and Intelligence,” Alan Turing addresses the question “Can machines think?” and argues that the question itself is too vague to be useful. Instead of trying to define “thinking,” Turing proposes a more practical test based on observable behavior, which he calls the imitation game. In this game, a human interrogator communicates through text with both a human and a machine. If the interrogator cannot reliably distinguish the machine from the human, the machine can be said to exhibit intelligence.
In the early sections of the paper, Turing explains the structure of digital computers and emphasizes that they operate by following discrete, well-defined rules. He argues that such machines are not fundamentally limited to simple calculations; in principle, they can simulate any process that can be described algorithmically. This leads to the claim that human reasoning itself may be representable as computation.
Turing also anticipates and responds to several common objections to machine intelligence. These include religious arguments (that thinking requires a soul), mathematical objections based on Gödel’s incompleteness theorem, arguments from consciousness and emotion, and claims that machines can only do what they are explicitly programmed to do. Turing counters these by suggesting that human intelligence is also constrained, that learning machines could modify their behavior, and that subjective experiences are not a reliable basis for judging intelligence.
Overall, Turing concludes that the question of whether machines can think should be replaced by empirical investigation: observing how machines behave as they become more sophisticated. He predicts that, over time, machines will become capable of performing well enough in the imitation game that objections to machine intelligence will seem increasingly irrelevant.
📝 HW2 – Due: Tue, Jan 20, 1:59 PM PT (before class)
HW2¶
Make a copy of HW2, answer the questions, and submit your responses as a PDF under HW2 on Gradescope.
Be creative! You may use an LLM to bounce ideas off of, but don’t outsource your thinking or creativity to it.
Even if you completed Activity 5 as part of the in-class worksheet, please make sure to submit a copy of that to HW2 to receive credit.
Grading
Graded for completion
Submission
Upload a PDF of your summary to Gradescope under
HW2
Late policy
Up to one week late accepted
- TURING, A. M. (1950). I.—COMPUTING MACHINERY AND INTELLIGENCE. Mind, LIX(236), 433–460. 10.1093/mind/lix.236.433