View source for Question about Economics (as a discipline)

{{pubdate|14|April|2026}}

Serious and honest question, and I'm not trying to diss on a different discipline: I don't have much experience reading papers in economics. And I saw that paper  "AI, Human Cognition and Knowledge Collapse", on a topic that I find very interesting, written by Acemoğlu et al. Now, Acemoğlu is a Nobel prize winner and MIT professor, so I would assume that paper is among the best economics would offer on that subject. 

The paper models several connections between different values mathematically, makes a number of assumptions, argue for these models and assumptions, and then draws the consequences of different scenarios based on those models and assumptions. 

But here's the thing that surprises me: yes, it argues for these models and assumptions, but it never actually validates them. It doesn't look at historical data, natural experiments, or actual experiments to test any of these. In the best case it may offer a citation to a paper supporting the argumentation, but that's it. 

Now  I do find a lot of the argumentation convincing, but that's probably biased by me finding the premise intriguing.

So, here's my question: is that how economics as a research discipline works? Is it because that's a "working paper", a kind of paper I'm not familiar with from computer science, and not a peer reviewed publication? Am I missing something? 

Besides all that, the paper is interesting and enjoyable to read, but I found myself sitting there wondering why I should believe any of the conclusions?

* [https://www.nber.org/papers/w34910 "AI, Human Cognition and Knowledge Collapse" by Daron Acemoğlu, Dingwen Kong, and Asuman E. Ozdaglar]

{{tag|Simia}}
<noinclude>{{simiapost|english}}</noinclude>