Phillip Hintikka Kieval

I am a PhD student in the Department of History and Philosophy of Science at the University of Cambridge where I am generously funded by the Gates Cambridge Scholarship.

In addition to my PhD, I also serve as a Student Fellow at the Leverhulme Centre for the Future of Intelligence.

I work on the history and philosophy of machine learning and applied statistics in science and public life. My dissertation articulates an account of deep learning models in science in terms of the practical articulation of stable design-rules for constructing reliable models tailored to achieve specific epistemic aims. Using the case of AlphaFold2 in structural biology, I argue that we can explain the effectiveness of deep learning across the sciences in terms of the implementation of domain-general mechanisms for enforcing transformational invariances that modellers choose based on background knowledge of physical regularities in a target system. This suggests that deep learning is continuous with a long history of applied mathematical techniques for constructing models that produce adequate predictions but do not necessarily aim to represent their targets.

I also have other projects in the philosophy of cognitive neuroscience, artificial intelligence, general philosophy of science, general epistemology, and social-political philosophy, especially where these areas intersect with aspects of machine learning.

Before Cambridge I was a student at the University of Houston where I earned my BA and MA in Philosophy.

I share my life with my brilliant wife and our two cats Winston and Lumi.