Better than they know themselves? Algorithms and subjectivity

The paper explores the widely circulated idea that algorithms will soon be able to know people “better than they know themselves.” I address this idea from two per- spectives. First I argue for the particular subjective qualities of experience and self- understanding issuing from our engagement with the world and the constitutive role of our reflexive relation to ourselves. These are not “known” by the algorithms. I then address our fundamental opacity to ourselves and the biased, partial, and lim- ited nature of human self-understanding. Our failure to know ourselves is however essential to our subjectivity and therefore, to know a subject in a perfect way that bypasses these limitations is actually not to know them. Taken together, both direc- tions show that while algorithmic knowledge of humans can be vast, and can outper- form their own knowledge, it remains foreign to their subjectivity and cannot be said to be better than self-understanding.