It looks like you're new here. If you want to get involved, click one of these buttons!
Following Jeremy's lead, I wanted to see what ChatGPT-4 would do when prompted to read the DOCTOR script through the lens of our Week 1 theme.
Prompt:
Acting as a critical code studies advisor to a scholar who wants to consider Weizenbaum's DOCTOR script in terms of Queer Theory: What parts of the code poem would you suggest that they consider, and why? For example you might suggest specific keywords, lines of code, or uses of syntax. Give concrete examples
Here was ChatGPT's reply:
Diving into Weizenbaum's DOCTOR through the lens of Queer Theory opens a fascinating dialogue between technology and the fluidity of identity, doesn't it? This perspective allows us to explore how the script navigates, and perhaps even subverts, traditional norms and binaries, much like Queer Theory itself.
Firstly, the very nature of ELIZA's script, with its ability to transform and reflect the user's input, can be seen as embodying the queering of conversation. It doesn't adhere strictly to a predefined narrative but instead fluidly adapts to the user's input, mirroring the non-linear, non-binary approach to identity and dialogue found in Queer Theory.
Consider the pronoun transformation functionality within the script:
(I = YOU
((0 YOU (* WANT NEED) 0)
(WHAT WOULD IT MEAN TO YOU IF YOU GOT 4)
...
(WHAT DOES WANTING 4 HAVE TO DO WITH THIS DISCUSSION))
This section blurs the lines between the self and the other, challenging the rigid binaries of identity. By switching pronouns, the script destabilizes the fixed subject positions and mirrors the fluidity of selfhood, echoing the queering of identity where the boundaries between the "I" and the "You" are not strictly defined.
Interestingly, one member of our team, Peggy Weil was also drawn to this portion of the script, though she took it in a different direction. What I like is that the LLM has taken up the notion of queering tied to selfhood but beyond sexuality.
Another intriguing aspect is the handling of ambiguity and non-specificity, which can be queer in its resistance to definitive answers or identities:
(NONE
((0)
(I AM NOT SURE I UNDERSTAND YOU FULLY)
...
(DO YOU FEEL STRONGLY ABOUT DISCUSSING SUCH THINGS)))
This part of the script allows for a multiplicity of interpretations and meanings, embracing the uncertainty and open-endedness celebrated in Queer Theory. It reflects the idea that identities and meanings are not fixed but are fluid and subject to change.
I found this section response to be more creative. It is using a very open or expansive sense of Queer Theory. And while I wouldn't follow its conclusions, I think that by taking up this line NONE, which is essentially a default situation for comments without recognized keywords, ChatGPT has found a notable moment of the code worth discussing.
The next example was not as compelling.
The script's neutral responses to inputs about dreams and desires also offer a queer reading. The dream sequences, for example:
(DREAM 3
((0)
(WHAT DOES THAT DREAM SUGGEST TO YOU)
...
(DON'T YOU BELIEVE THAT DREAM HAS SOMETHING TO DO WITH YOUR PROBLEM)
(NEWKEY)))
This segment doesn't assume a normative framework for interpreting dreams, desires, or aspirations. Instead, it invites the user to explore their own meanings and significances, resonating with Queer Theory's challenge to normative narratives and encouraging a personal exploration of desire that defies conventional categorizations.
This example felt a bit weak, but perhaps no more than that of a student just learning to try to apply Queer Theory to interpreting digital objects.
Lastly, the script's very impersonation of a therapist, a role traditionally associated with authority and normativity, can be seen as a subversive act. ELIZA queers the therapist-client dynamic by destabilizing the authority of the therapist, not through direct challenge, but through mimicry and performance, echoing Judith Butler's notion of gender performativity.
By advising a scholar to focus on these elements, you could illuminate how DOCTOR's script participates in a queer dialogue, not by explicitly addressing queer content, but by embodying the fluidity, performance, and subversion of norms that are central to Queer Theory. Through this lens, ELIZA becomes more than a technological curiosity; it becomes a participant in the ongoing conversation about identity, meaning, and the power dynamics of language and interaction.
Questions:
What do you get when you try using different LLMs to read the DOCTOR script?
What results do you get with different prompts?
Is it problematic that we're using AI on the script of someone who became so critical of AI as Weizenbuam?
What might ELIZA say about our exercise?
Are we co-reading with an LLM or disrupting the human activity of interpretation?
Comments
I gave your prompt to Le Chat and it said:
Minor issue: ELIZA/DOCTOR will never respond too "I am depressed" with "Why do you say you are depressed?" Also, here is how ELIZA/DOCTOR responds to "I am gay"
(Provocative responses?)
I then asked it for a line-by-line analysis of the DOCTOR script in terms of Queer Theory:
...and so on.