Lecturer in Comp Sci at Birkbeck, London. Natural language processing for social science and humanities. @[email protected] , pgnulty on bsky and threads
@TheStalwart It seems like they selected on the dependent variable to some extent: manually identifiying writing tics that Satoshi and Back shared, and then quantitatively confirming that those features overlapped with Back most from the list of candidates
Very happy that our review article "Dual-process theory and decision-making in large language models" has been published in Nature Reviews Psychology, with Oliver Brady, Lili Zhang, Tomas Ward and David McGovern @DP_McGovern (link below..)
Polars is a very fast DataFrame library for python with clean syntax and lazy evaluation - this course is a great guide to doing data analysis in python with polars.
Polars is the fastest data analysis tool on the 🌍
What's the fastest way to get going with Polars?
The Data Analysis with Polars course I'm launching today on udemy:
https://t.co/NmIT9Ax5Ly
I've priced the course to be accessible for all - it's just $15 at the moment.
@FelixHill84@spiantado @shimonedelman @DeepMind This paper has an interesting discussion about why and how expressions in text tend to correspond to facts in the real world https://t.co/SOn5NFMwf1
Tomorrow evening I'm talking in-person @BirkbeckUoL for #BEIWeek about opportunities for applying natural language processing methods to research questions in social science and the humanities. Details here: https://t.co/Bj9hobzwnS
@quadrismegistus@Ted_Underwood @Adam_Crymble @jeddobson @_akpiper I've stayed in "latest" mode and I still really like it. You have to be more selective in following but that's not a bad thing.
Newsletter! A lens on the magic of deliberation https://t.co/FsDAOTbriN let me tell you why I'm so excited about the work @G_Karadzhov has done, with myself and @vlachos_nlp
Between 2008/9-2020, there was a 9.5%⬇️in annual funding (real terms) for public HEIs in Ireland, a 3%⬆️ in academic staff, and , wait for it, a 31.7% ⬆️ in students. 😬
@janimarjanen If you don't mind the association with "essentially contested concepts", you can talk about "contested concepts". Describing the process as a "contest" maybe sounds a bit strange though (it's not a gameshow!). "dispute" is also a bit less dramatic than struggle
@adinamwilliams@dmimno There's a nice series of @LanguageLog posts about this: https://t.co/pQ2NGvyOvt I don't think they reach a clear answer, but it's not just a Google books thing.
@AaronRHanlon @quadrismegistus@Ted_Underwood @calebsmith203 Part of it is maybe that the subject of "pay" directs their own attention, while the other verbs refer to directing the attention of others. Interesting that it is economic when "paid" by the agent, but not when "gathered" or "drawn" like a natural resource