Discussion about this post

User's avatar
Alex's avatar

This does not demonstrate strong understanding of the philosophical arguments around AI early enough to make reading it a good bet for me. Is there a shot at buffing the intro up with something that makes me think the article isn't the following claims?:

1. AI can't really think; it only executes algortithms. Humans are special and do more than that.

2. AI can do superficial analysis, but not deep analysis, which humans can. Humans are often trained in deep analysis by starting with shallow analysis, but the easy path of using AI to cheat has broken this on-ramp.

Claim 2 is interesting and worth discussion, but claim 1 is better suited for an article with a technical background instead of just being repeatedly asserted. Claim 2 could use a summary-style inteoduction at the start to build reader trust, in my opinion.

Expand full comment
Eamonn Toland's avatar

AI's ability to mimic will only improve. Prose that might seem lifeless and banal today will become more sophisticated. We will eventually need AI tools to determine whether Humanities work is AI-generated or not.

AI has already had an enormous impact on STEM classes, with faculty reviewing Physics and Math homework to determine if it has been AI-generated. I have heard of students voluntarily retaking STEM classes they were acing because they felt they had become overly reliant on AI to complete their assignments, without understanding the fundamental principles of the course.

At the same time faculty are trying to encourage students to use AI tools appropriately, since they are rapidly changing the way we work. AI note taking at meetings already summarizes rambling discussions into salient points, often quite brilliantly, even if there are still occasional howlers. I have also heard of professors encouraging students to use AI tools to summarize lectures and notes into crib sheets that they can use to revise for exams.

I appreciate the cri de coeur on dumbing down, especially as it relates to the Humanities, but it's assertion-based. Schools and colleges urgently need more evidence on the impact of AI, not least as it relates to plagiarism and the effective use of revolutionary tools.

Expand full comment
3 more comments...

No posts