[46] Yulia Tsvetkov - Linguistic Knowledge in Data-Driven NLP
The Thesis Review
Paradigm Shifts in the Way We Build Models Today
In terms of technical modeling, I would like to see some paradigm shifts in the way we build models today. One example would be getting rid of sub-worlds, units that are not informed by what a world is or what a unit of meaning is and not transferable across languages. In terms of ethics, I don't know if it makes sense to build a big language model without ethical considerations, and then have ethical track papers that criticize and fix that model. This should not be disjoint research areas, right? It should be one of the considerations,. Like we preprocess our data in certain ways, and there should be considerations, what kind of problems we want to
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
