A few people have asked for the syllabus from my grad seminar on Generative AI for social science -- just posted it here:
statmodeling.stat.columbia.edu/2026/03/10/n...
Something I'm not getting here.... why use ATProto for this? How does this improve on, say, XML? Genuine question, and it it's answered in the docs someone, just point me at it.
Image with mushroom cloud and text: It's 11pm. Do you know what your expert system just inferred?
New essay on LLMs and brain rot:
"LLMs do not inevitably corrode thinking. They amplify whatever epistemic posture you bring..."
sbgeoaiphd.github.io/rotating_the...
lots of good points have already been made on using AI Agents for cheating (e.g. the latest Canvas-bot), it degrades learning, etc.
One additional thing I'd like to point out: if you use this stuff, you're not being clever, you're just an asshole.
to explain:
Very interesting... Where is this? I don't think I ever seen any official rubrics here for evaluating faculty.
Hmm. I see this from the opposite direction - working with these tools well for research is a form of natural language programming. I can bring my students closer to understanding that part of the code than I ever could in the one semester Python, one semester JavaScript constraints I had before.
Cool! Gotta try that.
Agentivity. With the verb "help" I want to know if the helper is capable of volitional action (a person, an organization) or not (an instrument, a circumstance). Seems like LLMs could be revolutionary for this kind of one-off project-specific annotation that would otherwise be way too expensive.
Trying to figure out what univ admin must believe in order to think the sector can survive paying for access to ChatGPT
I guess if you thought
• inference will rapidly commoditize
• new products will not emerge
• human guidance will remain the differentiator
it wouldn’t be self-destructive?
In some tests I'm doing, Opus beat Sonnet which beat Haiku. But gpt-oss-120b beat all of them! The differences were all small, though. I'll have to try having it correct its errors. I need to stop thinking of these as just slow, expensive taggers. They really are a different animal.
Huh, I haven't really considered keyness et al. But you could think of chi sq as a binomial regression where you want to know about the interaction term. Bayesianizing it isn't going to change the point estimate but might give you a different view of its uncertainty.
✋As it happens, I am at this very moment writing up an argument that only Bayesian stats make any sense for historical corpus linguistics (there's no population!)
I wonder how many papers in 2020 (and 1990) had at least one citation error.
Here's more recent project -- it includes Cherokee and gets Swahili right. No doubt there are still errors and omissions, they're unavoidable. They accept updates, though, so if you catch any you can let them know.
A rare derivational suffix that emerged from a frequent inflection is odd from a grammaticalization perspective. But it's a very natural development if seen in the context of the whole system.
It's a typologically unusual construction and a small corner of the Hungarian grammar, but it reflects large systemic changes in verbal inflection post-1600's: the shift away from periphrastic perfective marking and the loss the -ik paradigm.
Timeline (1500–2000) showing the development of Hungarian -hAtnék: from potential conditional verbs (1500), through divergence and reanalysis of hAt-n-ék → hAtnék (1600–1700), to first nominal uses (1800) and increased productivity (2000). Two bars note concurrent obsolescence of complex tenses and the ik paradigm.
link.springer.com/epdf/10.1007...
New paper out today in Morphology! Morphological change as systemically motivated bricolage: Hungarian impulsative constructions. We trace the development of Hungarian potential conditional verb forms (sírhatnék 'I could cry') to impulsative nouns (sírhatnék 'the urge to cry'). #linguistics
Definitely. But if AI makes it possible for people to make tools that scratch their own personal itches, "deployment" isn't as much of an issue. They're their only customers.
I mean.....I use the word enshittification and I also use claude code to build my own alternatives. As I'm convinced the tools I build actually work I'll release them. Don't really want people to use them though. I'd rather they build their own stuff. Anyway, maybe I'm unique? Don't think so, but...
I'm hiring! 2 x PhD positions to work at Uni Zurich with me on "Evolutionary Semantics" for an ERC-funded project. semantically curious MA grads from linguistics, cog psych, anthro encouraged to apply
evolvinglanguage.ch/studying-the...
a chicken behind the desk of an enormously outfitted computer station
don't worry guys I'm sure everything will be fine
This is my experience too - LLMs are often wrong, but thinking through their wrongness can be very clarifying
I've completed "Christmas Tree Farm" - Day 12 - Advent of Code 2025 #AdventOfCode adventofcode.com/2025/day/12
I've found it really helpful for TikZ diagrams
Don't forget Sinclair! Everyone forgets Sinclair. He's the one that put Firth's ideas into practice.
Dinsdag verschijnt mijn boek 'Woord voor woord: de verleden tijd van taal'. Daarin vind je onder andere 70 grote infographics.
Vandaag is het tijd voor preview nr. 2: 'Een suikerreis':
"What's the part number for the plastic plug thing on the thermostat housing of a 3.2l VR6 (Audi 8p)". It instantly came back WITH THE RIGHT ANSWER.
I, for one, welcome our new part-number-knowing overlords.
