With #Galactica and #ChatGPT I'm seeing people again getting excited about the prospect of using language models to "access knowledge" (i.e. instead of search engines). They are not fit for that purpose --- both because they are designed to just make shit up and because they don't support information literacy. Chirag Shah and I lay this out in detail in our CHIIR 2022 paper:

https://dl.acm.org/doi/10.1145/3498366.3505816

>>

#introduction time: I’m a final year doctoral researcher at the Melbourne Law School, working at the intersections of #lawandhumanities.

My current project brings my work as an activist lawyer in conversation with my experience as a theatre practitioner. I explore the jurisprudential learnings offered by creative practices of dissent against the colonial sodomy law in India.

I’m also interested in the conduct of legal pedagogy, exploring how we might bring queer-feminist-anti racist perspectives to bear on core law subjects.

My first book, Love and Reparation, was published by Seagull Books and distributed by University of Chicago Press in 2021. It draws upon court transcripts, affidavits (real and imagined), archival research, and personal memoir to re-stage the Section 377 litigation in India.

I'm also a commissioning editor over at the Critical Legal Thinking blog. Feel free to pitch me your ideas!

@academicchatter @phdlife

If you block that instance that's doing bad stuff EARLY:

1. You can actually *prevent* abuse rather than just feel bad that it's a thing that happens and wonder why nothing could be done about it

2. You can influence the other instance's policies if they see quick and decisive backlash

3. You can prevent an instance from becoming "too big to block" in the minds of others

But you have to actually block them at the FIRST sign of trouble, not after the damage is done