montgomery hill
doctormont.bsky.social
montgomery hill
@doctormont.bsky.social
linguistics, cultural/indigenous studies, computer science, artificial intelligence | 35 | indigenous, leftist, pragmaticist | assistant professor @ SUNY buffalo
I think it's fascinating that things like "a", "an", and other smaller discourse particles just get totally wiped out in LLM style analyses of language. Unless they don't? Correct me please!!
March 21, 2025 at 8:51 PM
I really appreciate how much natural light we get in my department’s new space…
March 14, 2025 at 1:02 PM
at least as the sole mechanism, so LLMs are participating in this human enterprise without the full scope of human cognitive capacity... it’s almost easier to imagine each AI model as a strange individual human with their own personality quirks w/r/t linguistic behaviors.
March 12, 2025 at 7:05 PM
So does anyone think LLMs are “doing language?” In one sense, yes, because what they are outputting is language. In another sense, I am not convinced there is a human out there that “does” language using transformers or diffusion …
March 12, 2025 at 7:05 PM
Y'know, since ideology shapes our understanding of at least our social reality, and language is the primary vehicle of ideology, I can't help but think about a fun conversation between Zizek and Whorf.
March 10, 2025 at 5:22 PM
splurging for the laptop with the touch screen was a great choice
March 7, 2025 at 8:03 PM
"To act intelligently and see intelligently become one" - C. S. Peirce
March 5, 2025 at 5:17 PM
The way you describe wolves howling in Mohawk, kontiwén:note, literally means they (konti) are standing (ote) the words (wenn) up. It is the same word for talking about a group of people speaking. #kanienkeha #linguistics #indigenous
March 3, 2025 at 4:02 PM
Dan Everett writing a biography on C. S. Peirce does not surprise me: Peirce's semiotic theory is probably the most significant possible challenger to Chomskyan thought about language. I would recommend reading Peirce. #cspeirce
March 3, 2025 at 3:56 PM
LMAO wow, ok — yes these two names are very helpful in understanding this table of contents
February 28, 2025 at 12:10 AM
February 26, 2025 at 2:20 AM
Today, I repaired my broken pair of glasses (pictured) today, and relieved myself of the headache my contacts were giving me. 🤓
February 25, 2025 at 11:44 PM
Anyone got a good youtube about the Minimalist program of linguistics? Or a podcast. ngl I'm into Lakoff/embodied cognition and Peircean semiotics. How is this one: www.youtube.com/watch?v=SJSz... #linguistics #minimalism #chomsky
Noam Chomsky, Genuine Explanation and the Strong Minimalist Thesis (MIT Linguistics Colloquium)
YouTube video by Abdessamia Bencheikh
www.youtube.com
February 25, 2025 at 4:48 PM
A community of settlers claiming to be Indigenous were recognized by the state of Vermont ten years ago, but it has come to light that the Vermont state legislature made a wrong call, and, imo, ought to rectify it: vtdigger.org/2025/02/20/a...
At contentious Statehouse panel, Abenaki leaders urge lawmakers to reconsider recognition of Vermont groups  - VTDigger
The event was hosted by Rep. Troy Headrick, I-Burlington, who plans to introduce a bill that would set up a task force to review past tribal recognition decisions in the state.
vtdigger.org
February 24, 2025 at 9:10 PM
Yeah, or if I’m shaking a dog’s hand / paw. I’m actually good calling it either of those in that situation, but not foot. Lol
February 24, 2025 at 2:31 PM
My current method is to grade them with the same rubric and just make sure to penalize AI misuse just as I would other kinds of factual or structural errors. I also let them know if they write something they don’t want to read, I don’t want to read it, either.
February 24, 2025 at 1:22 PM
Ok, so here’s a question. I have an undergrad course, I assign my students a 500-word response paper. They are allowed to use AI as long as they disclose AI. Should the AI and non-AI papers be graded with the same, or a different rubric?
February 24, 2025 at 1:22 PM
Depends on what they're up to.
February 24, 2025 at 12:17 AM
"Do not block the way of inquiry." - C. S. Peirce, 1899
February 23, 2025 at 6:33 PM
also a win: managed to switch my dad’s tv preferences from fox to msnbc, which had a good bit on about history and the war in Ukraine
February 23, 2025 at 4:31 PM
The presence of Ohio as a (meme) word in a significant amount of the populations’s vocabulary means that this same population is now highly cognizant of a word of Iroquoian origin - ohí:yo’ “beautiful river” #linguistics
February 23, 2025 at 4:29 PM
The only reason people want humanoid robots is because they want indentured servants, they have a god complex or both. A human is such a stupid design for a robot. However, most other animals (birds, snakes, fish) are cool designs for robots.
February 22, 2025 at 5:02 PM
Just used Zoom's AI summary for some meetings I had yesterday. I don't think I lost anything in using it. Mentally correcting the notes it sent after the meeting helped me consolidate my memory, even.
February 21, 2025 at 4:58 PM
February 20, 2025 at 6:45 PM
I'm glad the judge is making them listen to it even though they don't want to hear it.
Judge Reyes just became my favorite person.
February 19, 2025 at 8:05 PM