Abstract
AbstractAccording to the standard no miracles argument, science’s predictive success is best explained by the approximate truth of its theories. In contemporary science, however, machine learning systems, such as AlphaFold2, are also remarkably predictively successful. Thus, we might ask what best explains such successes. Might these AIs accurately represent critical aspects of their targets in the world? And if so, does a variant of the no miracles argument apply to these AIs? We argue for an affirmative answer to these questions. We conclude that if the standard no miracles argument is sound, an AI-specific no miracles argument is also sound.
Funder
Research Grants Council, University Grants Committee
Lingnan University
Publisher
Springer Science and Business Media LLC
Reference82 articles.
1. AlQuraishi, M. (2021). Protein-structure prediction revolutionized. Nature, 596(7873), 487–488.
2. Barnes, E. (2022). Prediction versus accommodation. In E. N. Zalta (ed.), Stanford encyclopedia of philosophy. Stanford University. https://plato.stanford.edu/archives/win2022/entries/prediction-accommodation/
3. Biddle, J. (2020). On predicting recidivism: Epistemic risk, tradeoffs, and values in machine learning. Canadian Journal of Philosophy, 52, 321–341.
4. Block, N. (1986). Advertisement for a semantics for psychology. Midwest Studies in Philosophy, 10, 615–678.
5. Boge, F. J. (2021). Two dimensions of opacity and the deep learning predicament. Minds and Machines, 32, 43–75.