In March 2026, OFAI will host a select group of researchers discussing the future of computational linguistics in the age of AI. The discussions will include but will not be restricted to questions such as: Is computational linguistics (CL) dead, or is there still a place for CL beside LLMs? Do Transformer-based LLMs or multi-modal models bring us closer to real natural language understanding (NLU), or do we need a different architecture for true progress? Do LLMs reduce everything to a linguistic problem? Etc. ...
Location:
Österreichisches Forschungsinstitut für Artificial Intelligence (OFAI) Freyung 6/6, 1010 Wien, Austria, 3rd floor (no elevator!)
Schedule:
-
Wed 04.03. afternoon: positioning & goals (starting with position statements from participants)
-
Thu 05.03.: brainstorming & discussions (based on key questions)
-
Fri 06.03.: summary, position statement, action points
More extensive list of key questions:
-
Is computational linguistics (CL) dead? Is there still a place for computational linguists beside LLMs?
-
What is CL as opposed to NLP? Should we today still talk about CL and if yes, what are the key aspects, what sets CL apart from natural language processing (NLP) and language technology (LT)? And what sets NLP/LT apart from AI?
-
What aspects are particularly CL-ish and why do we want to keep those (a) as research topics, (b) in training young researchers?
-
Should NLP focus on evaluation and data sets? Or can we still contribute to the development of NLP algorithms?
-
How can publicly funded research (the “leaking roofs“ model) compete with rapid industrial development (with a $500 billion budget)?
-
Do Transformer-based LLMs or multi-modal models bring us closer to real natural language understanding (NLU), e.g., where meaning evolves throughout the communicative interaction of two or more situated communicating agents. What else is relevant for NLU?
-
Will we need a different architecture for true progress?
-
How do we deal with LLMs? Trust them blindly? Try to make the best use of them? Reject them completely (→ Stephanie)? Attempt to evaluate them systematically?
-
Do LLMs reduce everything to a linguistic problem? There is neurobiological evidence that in human brains most tasks (e.g. logical reasoning, mathematics, theory of mind, …) are processed independently from language.
-
What impact will LLMs have on (corpus) linguistics? Do we still need human researchers or do we leave all interpretation to AI?
Related links:
SPP 2556 LaSTing “Robust Assessment & Safe Applicability of Language Modelling: Foundations for a New Field of Language Science & Technology“ addresses the relation between linguistics and AI – see https://www.lasting-spp.org/
