I was sad to hear that my former workplace, the Center for Theoretical Physics of Complex Systems, is winding down. It was such a great academic environment with time to think and ample opportunities to learn from colleagues and the regular seminars and international workshops. From the Center's last Scientific Report:
Outlook: The center counts 492 publications, a total of 12813 citations, and an h-index h = 56 on Google Scholar. Despite its tremendous success, a continuation with a new division headed by a new director could not be realized by IBS, which is a pity and raises other IBS related questions which are not part of the current report. As a result of the foreseeable retirement of the current director, the PCS is winding down by the end of 2025. Practically all members of the PCS quickly found or are successfully securing new positions in research institutes and universities worldwide. The successful concept of the PCS will continue to exist through its alumni who carry the message into the world. These include twenty three (23!) faculties worldwide, including eight (8!) in Korea, six (6!) in China, and five (5!) in India, but also in Singapore, Vietnam, Brazil, USA, and the Philippines.
It's a real shame, especially since support for similar theory-focused research centers is so limited. Short term grants promote "safe" topics rather than giving researchers the time and freedom to follow their curiosity and try new ideas.
Looking back, memorable moments at PCS include:
- A visit and seminar in 2018 by J. Michael Kosterlitz in which he recounted his unusual journey to his Nobel Prize-winning work, including the important role played by job rejections and rock climbing. We didn't record his talk, but what seems to be a similar version can be found here.
- Workshop weeks, particularly the ability to sit in on workshops beyond one's own areas of expertise and get a first-hand glimpse of how informal interactions differ between different fields. Sometimes the welcome reception and evening activities would wind down within an hour or so, other times they would continue into the early morning, prime time for forging new collaborations and hearing important gossip. This is also why online conferences are a poor substitute for in-person events.
- We went through a period where we were required (as a government institute) to have personal identifying information in all job applications be anonymised, to eliminate bias in their evaluation. Whoever came up with this didn't understand you cannot anonymise academic CVs - the publication list will inevitably give the name away!
---
Physical Review A saw a significant increase in submissions, including some LLM-written papers. When used properly, LLMs can be a great productivity enhancer, particularly for non-native English speakers. On the other hand, if one uses the LLM to "cheat" and write the paper entirely, it is really easy to spot. Some dead giveaways: formatting, em-dashes, fake references, overly wordy text that doesn't say much (or makes no sense at all).
For similar reasons it is easy to spot when a referee report (or student homework assignment) has been prepared using an LLM instead of real intelligence! During one of my classes this year, I was sad to see some students completing their hand-written humanities assignment by directly copying the output from ChatGPT. At the end of the day, tedious "homework" like unpaid reviews are not just a box to tick off, they are exercise for your mind. By looking carefully for flaws and inconsistencies in someone else's work, you are also developing the critical thinking skills that will improve your own writing and research. Don't short-change yourself by delegating to an LLM!
---
My own research is going at a slower pace. A lot of thinking time has been replaced with grant-writing. I hope to see some payoff for this substantial effort in 2026!
