michael
@_michaelginn
Followers
211
Following
73K
Media
59
Statuses
1K
PhD student at @CuLinguistics @BoulderNLP @lecslab. LLMs for rare languages, automata, synthetic data.
Boulder, CO
Joined November 2018
I’m mostly on bluesky nowadays: Also, check out my starter pack for NLP researchers:
0
0
0
Nailed it.
@skibidiquasar @inerati Between being lazy shits on social media, having AI do everything for you, and most notably people flooding into the field with no passion for it but who heard it's a way to get a 6 figure job, a larger percent of recent US CS grads are straight up ass. Get fucking good.
0
0
0
A CS degree doesn’t teach you how to be a software engineer. This worked fine back when CS students were already interested in the topic and doing side projects. This doesn't work now that the modal student is uninterested in anything except making money.
0
0
2
The average quality of junior devs has plummeted.
junior devs are fucked. the job market is really bad. outside of the tech twitter echo chamber, most freshers are not “cracked”. they don’t understand half of the words in junior dev job posts. the gap between what they know and what the companies require is HUGE.
0
0
0
first person to build an LLM for audio mixing will be a billionaire.
1
0
2
RT @ChengleiSi: Are AI scientists already better than human researchers?. We recruited 43 PhD students to spend 3 months executing research….
0
166
0
RT @colin_fraser: Many AI enthusiasts have just never thought carefully about like computer science or math or logic and I just can’t reall….
0
143
0
Find out in our new paper, which my colleague Ray Groshan will present at ACL!!.
arxiv.org
Data augmentation, a widely-employed technique for addressing data scarcity, involves generating synthetic data examples which are then used to augment available training data. Researchers have...
0
0
1
“Linguistically-motivated” techniques for data augmentation sounds good on paper, but is it worth the cost?.
1
0
2
“ChatGPT” is so awkward to say that gen z just calls it “chat”. “Codex” has been used for five different things. “o4-mini-high”. Seriously, what is going on with OpenAI naming??.
0
0
1
my heart goes out to the tragic victims of the hawk tuah cryptocurrency pump and dump scheme.
0
0
1
RT @BoulderNLP: 📢 Check out the lineup of papers our students will be showcasing at #EMNLP2024 in Miami next week! 🌴We'll be presenting new….
0
8
0
RT @BoulderNLP: 1️⃣ GlossLM: A Massively Multilingual Corpus and Pretrained Model for Interlinear Glossed Text. In-Person Poster Session B….
aclanthology.org
Michael Ginn, Lindia Tjuatja, Taiqi He, Enora Rice, Graham Neubig, Alexis Palmer, Lori Levin. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.
0
3
0
