# The responsibility to read in the Age of GenAI

In the Age of GenAI, code is paradoxically becoming both more and less prominent as a medium of communication. On the one hand, GenAI products such as ChatGPT, Cody and CoPilot have made it easier than ever before to write useful code. On the other hand, as more code is computer-generated, the ability of human coders to understand what they have written is under threat. This contradiction is acute in AI-Enhanced Humanities Research. Modern AI systems can turn in on themselves, helping the researcher write the code that ‘fine-tunes’ and ‘applies’ the model to the researcher’s question. The possibilities are immense, but so is the concern: How can we maintain research integrity in the Age of GenAI? What kind of expertise does the 21st-century humanist require? What should we do with the code that we generate, write and/or use?

In this workshop, the anticodians introduce the methods of Critical Code Studies (Marino 2020; Marino and Douglass 2023). Critical Code Studies combines two kinds of reading: technical reading of code for its functionality, and hermeneutic reading of code for its meaning. What unites these two sides of Critical Codes Studies is criticism. How does this code represent the world, the computer, the user, the programmer, and the reader? Why does it do so? The anticodians formed in 2024 to unpick these questions together. In this workshop, we invite conference attendees to join us for a half-day of slow, close, critical reading of a beautiful piece of source code.

# Format

The workshop will consist of a collective close reading of lis.py, an implementation of the Scheme programming language in Python. We will read the program line-by-line as a group, unpacking how each line works to build up a detailed picture of the software. Participants will be able to run lis.py on their own machine, supplementing their verbal reading with practical experimentation.

Our collective close reading will be interspersed with 5-minute lightning talks from the anticodians, lodging our close reading into broader contexts. Through these lightning talks, and free discussion throughout the collective close reading, participants will develop a toolbox for code critique, enabling them to critique software—including LLMs and their outputs—at a deeper level.

# Lightning Talks

Michael Falk | Lambda the Ultimate: Scheme and its ideologies

Kath Bode | ? The materiality of code ? Teacher or CoPilot: teaching programmatic reasoning today ?

Emily Fitzgerald | Coding Without Context?: Using ChatGPT to learn by doing

Leah Henrikson | Revisiting Computer Authorship: A Longitudinal Perspective on Social Perceptions of Computer-Generated Text Authorship

Dylan Chng | Ambiguous Poiesis, or Coding as Contingent Creation