Novice

PREDAVANJE PROF. DR. RALF-A LAEMMEL-A IZ UNIVERZE KOBLENZ

Datum objave: 3. september 2023 | Datum poteka: 12. september 2023

Novica je potekla!

Prof. dr. Ralf Laemmel iz Univerze Koblenz, Nemčija, bo imel v torek 12. 9. 2023 ob 11 uri v predavalnici Beta vabljeno predavanje z naslovom "The Paradigm Shift in SE & PL Research & Teaching due to LLMs". Več informacij na spodnji povezavi. Vljudno vabljeni!


Title: The Paradigm Shift in SE & PL Research & Teaching due to LLMs,

Abstract: AI has slowly (?) made it into the reality of software engineering (SE) and programming language (PL) research and teaching over the last 10-20 years. For example, in research, think of grammar inference and pretty-printing models in PL; think of bug-prediction models and search-based software re-engineering in SE. However, the capabilities of (contemporary) large language models (LLMs) in terms of understanding (?) of software languages and related tasks of evaluation, translation, summarization, etc. may be interpreted as reaching a new level of AI-related qualities in SE and PL research and teaching -- to the extent of challenging the status quo. In this meeting, let’s put on the SE/PL hat and discuss a few related questions such as these: i) How can stereotypical SE/PL "tasks" be differently approached in the view of contemporary LLMs and other AIs? ii) How do SE/PL courses need to be adjusted to appropriately develop student skills in the view of contemporary LLMs and other AIs? iii) How does SE/PL research need to be adjusted in the view of contemporary LLMs and other AIs? iv) How do SE/PL symbolic computation approaches of parsing, type checking, refactoring, etc. need to be reevaluated in the view of impressive (?) and contemporary LLM/AI capabilities -- does ML make symbolic computations redundant (to what extent)? v) How to usefully inject more classic SE/PL strongholds into LLM architectures so that we get a win-win (rather than hallucinating AIs and the lack of using strong symbolic computation capabilities)? vi) How to — in terms of (AI) ethics and policy —  redefine priorities or the value system in SE/PL research and teaching so that we — as a community — can make a useful contribution to a stable society in the near future?

The speaker: Prof. Dr. Ralf Lämmel, University of Koblenz, Germany
Joint work: Susanne Göbel, University of Koblenz, Germany

About the speaker: Ralf Lämmel held positions at University of Rostock, CWI & VU Amsterdam, Microsoft USA, University of Koblenz, Facebook London  as well as various appointments during sabbaticals. He is a W3 professor in CS and dean of the CS faculty at the youngest university of Germany — University of Koblenz. He is professor in software languages and his research and teaching interests relate to the intersection of the areas of software engineering and programming languages with particular interest in executable language definitions, software analysis and transformation, as well as AI applications in these areas. He is the award-winning author of the "Software Languages Book". He is a co-founder of the SLE conference and the GTTSE summer school series.

Links:
https://dblp.org/pid/l/RalfLammel.html
https://scholar.google.com/citations?user=rlhpWRQAAAAJ&hl=en