Ian Mulvany

May 31, 2025

Interesting links - 31st May 2025

I have continued to collect interesting links but noticed that I've not posted out any since the start of March, so I'm going to work through these over the next few weeks. This is the first batch. 



Taming LLMs
found on 2025–03–02




In brief, the sharp decrease in marginal utility of accolades, the voluntary nature of peer review, and the unexpected properties of accolade awards are the potential mechanisms that generate the negative effect. The academic community should reassess the existing incentive strategies for reviewers.

Peer review awards don’t work. (Discounts on APCs do)
found on 2025–03–02




Within an overarching AI governance policy, BMJ Group mandates authors who are submitting scholarly work to disclose and describe their use of AI tools, the reasons for using the AI, and the tasks completed by the AI. Objectives: To assess the frequency authors disclose using AI in manuscripts submitted to BMJ Group journals, the types of AI tools used, the reasons for its use, and the tasks it is used for.
found on 2025–03–03




This overview takes you through the key milestones and models that have made BMW one of the world’s most prestigious brands.

A really delightful site showcasing the history of BMW cars.
found on 2025–03–06




Mistral OCR is an Optical Character Recognition API that sets a new standard in document understanding. Unlike other models, Mistral OCR comprehends each element of documents–media, text, tables, equations–with unprecedented accuracy and cognition. It takes images and PDFs as input and extracts content in an ordered interleaved text and images.

There are many workflows in scholarly publishing where this kind of technology could help, have they become easier to implement now?
found on 2025–03–06




I am now re-examining some of the other things I often thought were too complex or too far above my abilities for hobby projects, and, so long as I can afford them, give them a go. It is an exciting time to be alive.

I think there are two dimensions of scale that AI in its current form gives us - 1) individual capability - 2) horizontal scale out. I really feel we are not thinking big enough yet.
found on 2025–03–10


About Ian Mulvany

Hi, I'm Ian - I work on academic publishing systems. You can find out more about me at mulvany.net. I'm always interested in engaging with folk on these topics, if you have made your way here don't hesitate to reach out if there is anything you want to share, discuss, or ask for help with!