jenslehmann, 11 months ago to random German

Some personal notes on topics raised at the #ACL2023NLP panel on LLMs:

• NLP and Computational Linguists diverging
• Vertical applications becoming horizontal layers
• Role of open and closed source LLMs depending on usage
• Environmental concerns relevant
• Malicious actors using LLMs and the need for regulations
• NLP is becoming the natural science studying how LLMs work
• Academic research and research at tech companies enabling diversity (in academia) and scale (in industry)

emilymbender, 11 months ago to random

#ACL2023NLP panel on LLMs: Dan Klein expresses enthusiasm for LLMs as representations of corpora for corpus linguistics. @meg adds only if the data is documented. 💯

emilymbender, 11 months ago to ArtificialIntelligence

Hey #NLP—we all know better than to use NLP tools that we can't assess, right? These folks are advertising on the #acl2023nlp Rocket.chat, but give 0 info about how their system was built or evaluated.

The link for http://generalizable.xyz goes to a page that literally just consists of their logo.

In what sense is this GPT? What training data did they use? What test data? What evaluation metric?

But more to the pt: We're here to learn from each other, not to read synthetic versions of papers.