Talks & press
Selected invited talks and media coverage of the work on AI and language.
Recent invited talks & public lectures
- Words From the Machine: Why AI Chooses the Words It Does (public lecture with Zina Ward, Leon County Main Library, Tallahassee, May 2026)
- The Human Side of Robo-Talk: Tracing Lexical Quirks from Model Training to Human Usage (University of Oxford, Nov 2025; invited)
- Fully Automated Identification of Lexical Alignment and Preference-Stage Shifts (LREC 2026, Palma de Mallorca)
- Why Does ChatGPT “Delve” So Much? (COLING 2025, Abu Dhabi)
- Word Overuse and Alignment in Large Language Models (BIAS 2025 Workshop, ECML-PKDD, Porto)
In the news
The work on AI and language has been covered in more than twenty countries and fourteen languages.
For interviews or comment on AI and language, do not hesitate to email me.
Selected coverage
- The New York Times
- The Economist
- The Guardian
- The Washington Post (AI & language)
- Science and Technology Daily (CN)
- Chosun Ilbo (KR)
- Newsweek
- Newsweek Japan
- El País
- The Telegraph: consultants
- Forbes
- PAP (PL)
- WIRED en Español
- Fast Company
- Rzeczpospolita (PL)
- ZDNet
- Olhar Digital (BR)
- TechRadar
- Presse-citron (FR)
- Science Friday
- Dataconomy (DE)
- Reuters Institute (Oxford)
- Playtech (RO)
- TechInsider (RU)
Full list of outlets
- The Washington Post (Romance vs Germanic)
- The Telegraph: MPs
- Digital Trends
- Daily Mail
- Yahoo News
- TechXplore
- Cybernews
- Minnesota Star Tribune
- FSU News (Aug 2025)
- FSU News (Feb 2025)
- Digital Trends (ES)
- TechRadar (FR)
- Teoria Digital (BR)
- Gazeta do Leste (BR)
- Fast Company México
- Panorama del San Jorge (AR)
- Radio3 Cadena Patagonia (AR)
- El Teatro Digital
- TechEDT
- iMEdD Lab (GR)
- Kikar HaShabbat (IL)
- Głos (CZ)
- iGate (UA)
- Muhalif (TR)
- Informiran (BG)
- Robotitus (PE)
- DigitalPortal (SK)
- Newsis (KR)
- Polskie Radio (PL)
Expert commentary
Where outlets have sought my commentary on research in AI and language.
- The Economist: the Paper trails column How to spot AI writing, written following an exchange with the journalist, credits by name the human-feedback account of LLM word preferences: models pick up what readers find impressive and drop what they do not. Republished under licence in Portuguese by Folha de S.Paulo and in South Africa by BizNews. (July 2026)
- The New York Times: Vauhini Vara’s essay A.I. Loves the Em Dash. What’s a Human Writer to Do? includes a brief discussion of data-informed observations on em dash usage by AI systems. (June 2026)
- NBC News: invited, as an independent expert, to comment on Abdulhai et al.’s How LLMs Distort Our Written Language. (March 2026)