Source Posts on X
a. an ai system is more than a model and increasingly includes a variety of technical and human dependencies (e.g., tools, apis, internal policymakers). the diagram of a social media platform [Lum and Lazovich 2023] can easily be draped over an ai system. 2/6
2:17 PM · Jul 14, 2026b. ai models are going to be increasingly adaptive at individual and population levels (e.g., personalization, implicit feedback, for an individual or in aggregate). this means that, like recommender systems, pinning exactly what a "model" is at any one point is fraught. 3/6
2:17 PM · Jul 14, 2026standards bodies are good, but should go beyond benchmarks or laboratory testing. given the current state of ai evaluation, it's worth mentioning a few reasons why benchmarks are brittle... 1/6
2:17 PM · Jul 14, 2026Combined views
1.1K
3 posts, first seen 4h ago