Contra: The Machines are Fine

This is a short comment on The Machines are Fine by M. Karamanis. In it he primarily discusses LLMs and their role in science; how he is against a complete and blind adoption of them and how they will ultimately lead to the lack of actual understanding happening.

Unlike medicine, where a cure for Alzheimer's would be invaluable regardless of whether a human or an AI discovered it, astrophysics has no clinical output. The results, in a strict practical sense, don't matter.

This is, in a sense, true. But it is to me a touch too anthropocentric. I care not one deut about where knowledge comes from. The codification of the natural world is of intrinsic value, if only from a Last Question type argument. In a sense the argument comes from a feeling of lament about ones own role; but the wish for ones own job to be necessary ad infinitum is a bit small minded, even narcissistic - Does a general not dream that his own prowess is enough to end all war? Does the doctor not wish to cure the last disease. Is the best house not one that need never be repaired. The craftsman who cares about the work itself, not the validation he receives through its completion, wishes for a job that is done, and done well.

I am very willing to concede a point, which is that the death of a craft is in part a sad thing for individuals, and that the amount of understanding is going down, and will probably concentrate ever more into a small group of cognitive Elites. There is the idea that it is the friction in basic low level action, in the carrying out of exercises, that promotes learning the motion. This admiration of grunt work is, like it or not, the most replicated finding in pedagogy.

Deliberate practice requires effort and is not inherently enjoyable. Individuals are motivated to practice because practice improves performance.
~KA Ericsson