Tutorial #5 – research paper

17/06/26

Research paper tutorial with Janet

Things discussed:

• This post is a bit belated, to say the least. I completely forgot that I didn’t write this meeting up. Luckily, I made some good notes!
• Janet was very positive about my research paper draft. We talked through the proposed structure and agreed that it feels like a solid basis to build on.
• We also discussed the merits of using Refik Anadol and Martha Rosler as the main examples, and Janet seemed to think that this comparison had potential.
• One thing Janet was keen for me to pursue was Anadol’s relationship with companies such as Palantir and Google, especially in relation to his claims around creating an ‘ethical’ AI. This feels like an important thread. I chose Anadol partly because he is so unabashedly working with AI and with large corporations, but I do not want to set him up too simplistically as the ‘bad guy’. I am keen to write something balanced.
• We talked about the idea of ‘ethical AI’ and whether this is even possible. Even if Anadol is now using ‘ethically sourced’ datasets, there is still the question of how the underlying machine learning systems were trained in the first place. Derrida feels useful here. How do you draw a line in the sand and say, ‘this is the starting point’? Where is the original source? Can there ever really be one?
• We also discussed Anadol’s claims that he is “democratizing” and “liberating” data. What does this actually mean? Who is the data being liberated from, and who is it being liberated for? We also talked about the visual sameness of a lot of his outputs. Why do they all seem to look so similar? Is this an aesthetic choice, or is it partly the averaging effect of AI?
• Steyerl feels useful for thinking about the difficulty of pinning down singular ownership when it comes to contemporary images, especially when images are constantly circulating, being reused, compressed, transformed, and recontextualised.
• Debord also seems like a good comparative lever. I was slightly concerned that bringing him in as well might be too much, or that I might be trying to shoehorn him into the paper, but Janet was enthusiastic about his inclusion.

Next steps:

• Continue researching and to gather stronger references. The structure seems good at this point, but the argument needs to be supported more carefully. 
• Expand the draft to around 2500 words.

This is the current draft. I am putting it here because it will be replaced by the next version on the allocated page of the journal: