Join 📚 Michael's Highlights

A batch of the best highlights from what 🤓 Michael's read, .

The choice of an ML algorithm, the data fed into it to train it, the point at which it is considered adequately trained to be released, how that point is detected by testing, and whether that testing is ongoing if the learning continues during the system’s operation—all of these things are design decisions that not only must be made but also can easily be documented.

The Oxford Handbook of Ethics of AI

Markus D. Dubber, Frank Pasquale, and Sunit Das

While we can publish information regarding data, not all participants can grasp it with the same breadth and depth. The communication of an insight is more difficult. Clues can be given, but the light that brings understanding must come from inside; it must be generated by the person’s intelligence, intellectus agens. No one can substitute for another.

Discernment

Ladislas Orsy, SJ

Domine non sum dignus.

The Critical Calling

Richard A. McCormick and Lisa Sowle Cahill

...catch up on these, and many more highlights