Wurdaan Kashif
AI being trained on significant sets of data, much of which is taken and used without the consent or knowledge of the parties involved, is not particularly surprising to anyone (Villegas et al.). Further still, the reality that our data is constantly being sold and mined is also not unknown knowledge. I suspect that most people, even those who may understand the implications of what a ‘cookie’ is and what could possibly be hidden in a ‘terms and conditions’ page, click ‘accept all’ without much thought, or maybe even the cheeky ‘only essential’ if they’re feeling particularly privacy conscious that day. It not just appeals to my own nature, I should point to; in a study with a fake privacy agreement, researchers found 98% of Americans signed away their rights to their first born child (Cruz).
This is unsurprising, and it’s to be expected. The individual barriers are significant enough.
Firstly, in a world where significant amounts of populaces are illiterate, many just lack the literal comprehensive capacity to understand what is actually being taken from them (National Literary Institute). Secondly, and let it not be underplayed the significance of this reason, it is far more convenient and easy to just accept. An individual may join a site or sign some conditions because they want to be able to talk to their friends and family (Jimenez). They could be wanting to track and record personal medical history, trying to get a job, hoping to share art, all for the numerous reasons one might be inclined to do such things (‘Meta Eavesdropping Case‘, Jencer, Chen). However, it is equally important to recognise less societally ‘justified’ examples too. Whether it be playing a video game, trying to buy yourself something nice, or even watching pornography, these acts of menial pleasure are things people should be able to access without having to negotiate a contract over their data (‘Data Analytics Games‘, Kelion, Yount). When nearly every part of online existence is ubiquitously covered in tripwires, it only takes you to trip over one before you start ignoring their existence altogether. Pacifism is the design strategy.
The final reason I want to offer is the one my project aims to capture. That is, the psychological comprehension of what is actually happening. Humans are beings of cognitive dissonance, being able to hold and operate multiple facts of reality simultaneously (Perlovsky). Humans are also terrible at visualising, making very different decisions when they are making decisions based on the abstract-ification of understanding, compared to when they are approached with the same decision in reality (Bostyn). This problem of visualisation has a number of consequences when we consider the impacts of our actions, but this problem extends to ourselves and sometimes to the data collectors themselves too (Nijsse). We may, in some abstract sense, know what data is, and be familiar with how it is used in training, marketing, surveillance (etc). But it is very easy to convince ourselves that, even if this process is occurring, we are free from the consequences of the actions, or, in some way, not harmed by it.
The ‘datafarmer’ character turns what is previously digital into the personal. For as much as I attempt to frame the experience behind comedy (and I do sincerely hope you found it enjoyable), the absurdity that I am aiming to capture is the everyday theft that occurs to you, me, and nearly everyone we know of. What is the fundamental difference between a man taking photos, measurements, notes, of every move you take, and how companies collect our data?
There are two potential objections to this. The first is an appeal to essentialism- these are different scenarios, the real and the digital are fundamentally different. This argument is nonsensical. Just as there is no difference between dropping a bomb yourself or using a drone to do the task, the separation via the digital tool does not change the actions the tool is being used to perform (Knight). The second is an appeal to consent. We still did fundamentally choose to give up our data to these companies. To defend this objection, you would have to ignore the arguments I’ve provided as to how I believe we cannot call this a legitimate choice. You would even have to ignore the larger issue that, in many instances, AI companies would take data from sites even when the initial consumers of that site had no knowledge or compensation that it was being used in that way.
However, let’s say that the T&Cs did contain information that the website would share the data with AI companies. Let’s even assume that the signing of the contract is all that is needed to prove consent to use and sell data. This objection still does not show that there is a difference between the actions of my data farmer and the data scraping of these companies. In fact, it serves to legitimise a much broader criticism, that we (the individuals) have quite literally next to no control over any of our information anymore. If signing a T&C allows a company to use the likeness from your photos to better identify protestors in a crowd, is this digital social contract we should be supporting? If there really was a physical man present with you at every part of the day, listening to you speak, following you about, recording what you read and watched, how would that genuinely make you feel? The absurdity that visualising that reality creates is a defence against the feeling of discomfort it naturally provides.
This is what my project aims to do. This discomfort should not be run away from. It is that discomfort that is capturing how comfortable, or not, you really are with the current state of affairs. What is being captured and not captured? What is known about me? Why me? How is that data being used? Appreciating those questions begins to elucidate both the importance of our individual data sets and the degree of information that these companies wield and know about us. The end of my film has no real resolution, with the data being mailed off to OpenAI’s headquarters. What then, are they actually doing with it all? What are the actual consequences? There are a number of sources, accounts, and current examples of what the consequences may well be. I’d propose a different question. You’re an AI company; you need to make some money; you’ve got all the information in the world at your disposal, whenever and however you’d like. What can you think of?
Bibliography
Cascone, Sarah. ‘A Class Action Lawsuit Has Been Filed Against AI Generators DeviantArt, Midjourney, and Stable Diffusion.’ Artnet News, 17 Jan. 2023, news.artnet.com/art-world/class-action-lawsuit-ai-generators-deviantart-midjourney-stable-diffusion-2246770.
Chen, Angela. ‘LinkedIn is Training its AI on Your Data Before Updating its Terms of Service.’ Vice, 18 Sept. 2024, www.vice.com/en/article/linkedin-training-ai-on-member-profiles/.
Cruz, Brett ‘American Digital Privacy Knowledge.’ Security.org, 1 Apr. 2025 www.security.org/digital-security/american-digital-privacy-knowledge/.
‘Data Analytics Games: Gamifying Insights.’ CastorDoc, www.castordoc.com/ai-strategy/data-analytics-games-gamifying-insights.
Goldhill, Olivia. ‘What a Real-Life Trolley Problem Reveals About Morality.’ Psyche, 10 May. 2021, psyche.co/ideas/what-a-real-life-trolley-problem-reveals-about-morality.
Hao, Karen. Empire of AI : Inside the Reckless Race for Total Domination. Penguin Books, 2025
Hill, Kash. ‘Meta’s AI Scraping Policy: What to Know’ The New York Times, 11 Jun. 2024, www.nytimes.com/article/meta-ai-scraping-policy.html.
Kelion, Leo. ‘Amazon: The Truth About Your Data.’ BBC News, 2024, https://www.bbc.co.uk/news/extra/CLQYZENMBI/amazon-data.
Levy, Steven. ‘AI Lab Scout AI Is Using AI Agents to Blow Things Up.’ Wired, 18 Jun. 2024, www.wired.com/story/ai-lab-scout-ai-is-using-ai-agents-to-blow-things-up/.
National Literacy Institute. ‘2024-2025 Literacy Statistics.’ The National Literacy Institute, 2024, www.thenationalliteracyinstitute.com/2024-2025-literacy-statistics.
Nijhawan, L. P., et al. ‘Informed Consent: Issues and Challenges.’ Journal of Advanced Pharmaceutical Technology & Research, vol. 4, no. 3, 2013, pp. 134-40, www.ncbi.nlm.nih.gov/pmc/articles/PMC3622034/.
Nijsse, H. Privacy and Data Protection in Software Services. VLeBooks, 2024, www.vlebooks.com/Product/Index/4930206.
Villegas, Carol C., et al. ‘The ‘AI Revolution’ Comes With Data Privacy Risks: What Consumers Should Know.’ New York Law Journal, May 2024, https://www.law.com/newyorklawjournal/2024/05/17/the-ai-revolution-comes-with-data-privacy-risks-what-consumers-should-know/?slreturn=20260404092149.
Yount, Ashton. ‘When You Watch Online Porn, Who Is Watching You?’ Annenberg School for Communication, University of Pennsylvania, 17 Jul. 2019, www.asc.upenn.edu/news-events/news/when-you-watch-online-porn-who-watching-you.
Cite this page:
Kashif, Wurdaan. 'I Became A Data Farmer For A Day'. Cream of the Slop. version 1.0, Digital Humanities for Literary Studies 2025-26, University of Edinburgh, 10 Apr. 2026, https://blogs.ed.ac.uk/dh2025-26/.
