I am going to try something new with this blog, instead of updating it infrequently with whatever happens to catch my interest when I have the time. Over the past few months, a number of topics which long preoccupied my thoughts — secularization, political thought, data science/AI/Machine Learning, “internet culture” (i.e. Twitter), the philosophy of science, etc. — have coalesced into the idea of a somewhat coherent writing project.
The plan is to write a series of posts which attempt to place data/AI in a broader historical and intellectual context. People often say that Silicon Valley types need to learn more about the humanities, in the hope that this would open their souls to the world and make them behave more ethically. Setting aside the question of whether this would work for the moment (it would not), my project is not one of educating tech workers about philosophy. Nor is it, strictly speaking, a critique of data/AI using the concepts provided by the academic humanities.
What I am interested in, rather, is the fact that the discourse around data takes up a number of problems and debates that have, in the past, been discussed extensively within other intellectual traditions. This is not a bad thing; it normal that public discourse would confront the big, perennial questions of human existence, which philosophers et al. have traditionally argued over. My goal is to excavate these connections in the hope that they can suggest more productive ways of addressing the dilemmas raised by Data. (From here on out I will employ the capitalized “Data” to capture the whole gamut of related technologies, from internet cookies to SQL databases, Machine Learning algorithms, Artificial Intelligences, autonomous vehicles, etc.)
When I talk about the “discourse around Data,” I mean it very broadly — I will discuss not only books and articles on the topic, but also marketing copy, popular films, and of course, tweets. To illustrate what sorts of intellectual traditions I hope to cover, here are two examples:
Theology: As I have written previously, popular culture often depicts data mastery as granting god-like powers to its practitioners. This presents an entryway for thinking through how we conceive of the knowledge and control that our digital footprints concede to others, as well as common feelings of outrage, disgust, and fear toward tech titans like Mark Zuckerberg. Their ambitions seem to violate the natural order by usurping attributes of God — omniscience, omnipotence, immortality. Perennial philosophical debates, for example about free will and determinism, have traditionally been conducted in a theological register. Now they are present in conversations about data privacy, social media misinformation, really all over the place. My plan is to unpack how secularized theological concepts play a crucial role in shaping our concept of Data.
Epistemology: I have also written previously on this blog about the epistemological problems that are at play in data-based technologies. The question ultimately concerns the kind of knowledge that data can provide. Data is ultimately an abstraction, a digital footprint stored on a server. Combined with sophisticated statistical operations, these records of interactions can predict the future — based on the assumption that the past behaviors will continue. Although an algorithm can be wrong in its specific predictions, at scale it should make mistakes in a predictable fashion. While these insights may seem rote to a data scientist, or indeed anyone who knows a bit about statistics, I plan to explore these questions about certainty, doubt, predictability, and so on. Whether or not you understand something like Bayesian inference, the increasing role of statistical methods in all sorts of decision-making suggests that the underlying epistemological assumptions bear interrogation.
I will also discuss disciplines including sociology, ethics, history, political philosophy, phenomenology, political economy, and others, to the extent that my limited expertise allows. Additionally, several themes will arise repeatedly in the course of the blog entries. These are loci where Data and intellectual history intersect in particularly striking ways. Some of these are:
- Scale
- Complexity
- Emergence
- Agency
- Alienation
- Legibility
- Commodification
- Abstraction
- Simulation
- “Visualization”
- Taxonomy
- Identity
And so on. I hope to explore these themes over the next few months in the informal context of this blog. If you have any thoughts or comments on these issues, I am always happy to discuss them here in the comments section or on Twitter. I would love to hear from people who know more about both Data technologies and the intellectual histories under discussion.
HUGE REALITY 