Simon Münker - Notes

ACL 2026 - San Diego

Simon Münker — Thu, 02 Jul 2026 00:00:00 GMT

Hundreds of jet-lagged academics, the bleeding edge of artificial intelligence, huddled beneath a giant vinyl banner, competing for corporate giveaways and lukewarm hotel catering.

Paying top dollar to watch a hippo lie completely motionless in murky water for forty-five minutes. Watching a massive mammal exist, floating inertly while staring into space.

Thousands of academics are on the deck of a decommissioned aircraft carrier because nothing naturally complements natural language processing quite like cold ocean wind and military hardware.

A vintage silver Mercedes next to puddles and tagged concrete walls in a dreary back alley felt like a fitting metaphor for the entire trip: a slick, expensive, parked right on top of gritty infrastructure.

EACL 2026 - Rabat

Simon Münker — Fri, 24 Apr 2026 00:00:00 GMT

The EACL 2026 convention center, more like an ancient fortress, is gatekept by a security guard.

A fishing boat returned at dawn, unloading its cargo in Rabat harbour, opposite residential buildings standing at the coastline.

A flashing, retina-searing geometric LED wall blasting neon light inside the tent where the social event was located.

A lone figure fishing from a broken concrete block along the Atlantic coastline, miles away from the nearest server rack. Completely unbothered by AI and data centers.

HICSS 2026 - Maui

Simon Münker — Mon, 05 Jan 2026 00:00:00 GMT

My absolute sacrifice of academic life. Flying halfway across the globe to present a 15-minute slide deck in a tropical paradise. The registration desk was strategically placed right by the ocean breeze inside the Hilton resort.

Tourists lined the shore, angling their phones for a golden-hour selfie, while a kid in the surf stood judging our collective inability to enjoy a sunset without documenting it.

The official social event treated us to “authentic” Hawaiian culture, featuring talented dancers and musicians on stage. Local immersion on the manicured lawn of a luxury property that costs a quarter of a month’s research stipend per night.

Down the road stands the Old Lahaina Center. Spared by the wildfires, it now sits completely deserted, slowly rotting away in silence, with no open storefronts and empty lots surrounding it.

Introduction to Linguistics

Simon Münker — Thu, 01 Jan 2026 00:00:00 GMT

This lecture offers a broad introduction to linguistics, moving from the origins of human communication through to its modern-day collision with generative AI.

We open by asking what makes human language unique, comparing it against communication in bees, birds, and primates, before tracing the history of linguistic thought from ancient Greece through the Renaissance. From there, the course works systematically through the core levels of linguistic structure: phonology (sounds and phonemes), morphology (word formation and inflection), syntax (sentence structure, formal grammars, and dependency relations), semantics (meaning, compositionality, and semantic roles), and pragmatics (speech acts and conversational maxims).

The second half turns computational: corpora, word frequency analysis, and named-entity recognition lead into a hands-on case study of fine-tuning language models to imitate politicians on social media, including data collection, cleaning, training, and evaluation, warts and all. From there we examine how language behaves in digital spaces, from ICQ-era slang to hashtag grammar, before closing with a critical look at generative AI: the ethical costs of large-scale training, the “stochastic parrot” debate, and the linguistic fingerprints that still distinguish machine-generated text from human writing.

Students will annotate real text data as part of ongoing research, connecting classroom concepts directly to open questions in computational linguistics. All course content is available on StudIP via access as an enrolled Trier University student.

Research Case Studies

Simon Münker — Thu, 01 Jan 2026 00:00:00 GMT

Research Case Studies is a project-based module with students splitting into small groups of five to seven, and each group works quantitatively on a real research question for a full semester, from a feasibility pitch to a data-driven NLP pipeline to a poster presentation. Groups are self-organizing, with supervisors guiding rather than managing: setting expectations, checking in biweekly, and helping steer scope so the semester doesn’t dissolve into busywork on non-essential parts of the project.

This year I’m supervising a group on AI Psychometrics: Construct Validity and Biases in LLMs for Social Science. The aim is to investigate how well Large Language Models (LLMs) hold up as stand-ins for human respondents on established psychometric instruments, and to surface the systematic biases that show up when they’re pushed into that role. It’s a direct extension of the work I did throughout my Ph.D. on LLMs as human simulacra, and this time the goal is to get the students’ findings into shape for submission to a workshop, turning a semester project into an actual contribution to the field, not just a poster and a portfolio.

ACL 2025 - Vienna

Simon Münker — Sun, 27 Jul 2025 00:00:00 GMT

My first time at a major ACL venue. A gigantic convention center near the UN headquarters. ACL feels like something important. However, questions emerged about whether academic researchers are becoming “corporate drones” and whether industry is steering ACL research directions. Eduard Hovy, one of ACL’s founders, heavily criticized this trend and the publication of “LLM junk papers”: research that naively applies LLMs to demonstrate capabilities. Is what we are doing actually worth it?

Outside the conference, I explored Vienna and visited the Albertina Museum. A nice contrast to spending most of the day surrounded by papers, posters, and talks. Surprisingly, they had an after-party, which I joined at the O-Club. A completely different kind of conference experience than I envisioned.

At a ramen spot in a gentrifying quarter, a long queue stretched along the sidewalk. Naturally, we stood patiently in line, unwilling to settle for the second-best ramen restaurant in town.

Modern high-rises with tinted balconies overlook communal gardens and open walking paths where residents grow their own plants.

EMNLP 2025 - Suzhou

Simon Münker — Tue, 04 Nov 2025 00:00:00 GMT

Another conference, another excuse to travel across the globe to talk about language models with people who haven’t seen natural sunlight in months. EMNLP 2025 landed in Suzhou, a surprisingly charming mix of high-tech, ancient canals, and existential AI dread.

A massive bridge with the entire ceiling covered in LED panels displaying shifting animations and advertisements. Such modern. It feels like walking through Cyberpunk 77, but without bugs and meticulously cleaned.

A local religious community walks in slow circles, chanting to their gods. Watching them perform their ritual felt strangely like praying that the reviewers would show mercy on my latest submission. I should find community for that.

A traditional canal boat reimagined with bright LED strip lights and outboard motors, filled with tourists and floating through the Venice of China bathed in neon glow alongside dozens of fellow tourists.

Squeezed into an alleyway, hundreds of wooden plaques, a collection of written spells and wishes from locals and visitors praying for financial freedom, good health, and intellectual success.