(Un)rigorous AI

Imperfect, brief reflections about (un)rigorous AI work. About 500 words or less. These are tagged as either back-to-basics or current-issues. I love and, thus, probably overuse em dashes.

Back-to-basics: so what do evaluations do

TL;DR — Developing useful evaluations requires clarity about what exactly one is trying to learn about a phenomenon of interest.

Alexandra Olteanu
Aug 2

Back-to-basics: unobservable constructs and their ‘surplus meaning’

TL;DR — Because unobservable constructs hold ‘surplus meaning,’ your metric is not your construct.

Alexandra Olteanu
Jul 25

Back-to-basics: on poor conceptualizations in AI work

TL;DR — Poor conceptual foundations can severely undermine the credibility and reliability of knowledge claims. (And, no, your metric is not your construct.)

Alexandra Olteanu
Jul 20